[llvm] 4454152 - AMDGPU: Replace copy-to-mov-imm folding logic with class compat checks (#154501)
via llvm-commits
llvm-commits at lists.llvm.org
Tue Aug 26 07:41:38 PDT 2025
Author: Matt Arsenault
Date: 2025-08-26T23:41:35+09:00
New Revision: 445415219708f9539801018e03282049ca33e0e2
URL: https://github.com/llvm/llvm-project/commit/445415219708f9539801018e03282049ca33e0e2
DIFF: https://github.com/llvm/llvm-project/commit/445415219708f9539801018e03282049ca33e0e2.diff
LOG: AMDGPU: Replace copy-to-mov-imm folding logic with class compat checks (#154501)
This strengthens the check to ensure the new mov's source class
is compatible with the source register. This avoids using the register
sized based checks in getMovOpcode, which don't quite understand
AV superclasses correctly. As a side effect it also enables more folds
into true16 movs.
getMovOpcode should probably be deleted, or at least replaced
with class check based logic. In this particular case other
legality checks need to be mixed in with attempted IR changes,
so I didn't try to push all of that into the opcode selection.
Added:
Modified:
llvm/lib/Target/AMDGPU/SIFoldOperands.cpp
llvm/test/CodeGen/AMDGPU/br_cc.f16.ll
llvm/test/CodeGen/AMDGPU/call-argument-types.ll
llvm/test/CodeGen/AMDGPU/gfx-callable-argument-types.ll
llvm/test/CodeGen/AMDGPU/gfx-callable-return-types.ll
llvm/test/CodeGen/AMDGPU/imm16.ll
llvm/test/CodeGen/AMDGPU/llvm.amdgcn.interp.inreg.ll
Removed:
################################################################################
diff --git a/llvm/lib/Target/AMDGPU/SIFoldOperands.cpp b/llvm/lib/Target/AMDGPU/SIFoldOperands.cpp
index 962c276bc2123..d72af06ac566e 100644
--- a/llvm/lib/Target/AMDGPU/SIFoldOperands.cpp
+++ b/llvm/lib/Target/AMDGPU/SIFoldOperands.cpp
@@ -1248,6 +1248,7 @@ void SIFoldOperandsImpl::foldOperand(
if (FoldingImmLike && UseMI->isCopy()) {
Register DestReg = UseMI->getOperand(0).getReg();
Register SrcReg = UseMI->getOperand(1).getReg();
+ unsigned UseSubReg = UseMI->getOperand(1).getSubReg();
assert(SrcReg.isVirtual());
const TargetRegisterClass *SrcRC = MRI->getRegClass(SrcReg);
@@ -1278,44 +1279,60 @@ void SIFoldOperandsImpl::foldOperand(
DestRC = &AMDGPU::SGPR_32RegClass;
}
- // In order to fold immediates into copies, we need to change the
- // copy to a MOV.
+ // In order to fold immediates into copies, we need to change the copy to a
+ // MOV. Find a compatible mov instruction with the value.
+ for (unsigned MovOp :
+ {AMDGPU::S_MOV_B32, AMDGPU::V_MOV_B32_e32, AMDGPU::S_MOV_B64,
+ AMDGPU::V_MOV_B64_PSEUDO, AMDGPU::V_MOV_B16_t16_e64}) {
+ const MCInstrDesc &MovDesc = TII->get(MovOp);
+ assert(MovDesc.getNumDefs() > 0 && MovDesc.operands()[0].RegClass != -1);
+
+ const TargetRegisterClass *MovDstRC =
+ TRI->getRegClass(MovDesc.operands()[0].RegClass);
+
+ // Fold if the destination register class of the MOV instruction (ResRC)
+ // is a superclass of (or equal to) the destination register class of the
+ // COPY (DestRC). If this condition fails, folding would be illegal.
+ if (!DestRC->hasSuperClassEq(MovDstRC))
+ continue;
- unsigned MovOp = TII->getMovOpcode(DestRC);
- if (MovOp == AMDGPU::COPY)
- return;
+ const int SrcIdx = MovOp == AMDGPU::V_MOV_B16_t16_e64 ? 2 : 1;
+ const TargetRegisterClass *MovSrcRC =
+ TRI->getRegClass(MovDesc.operands()[SrcIdx].RegClass);
- // Fold if the destination register class of the MOV instruction (ResRC)
- // is a superclass of (or equal to) the destination register class of the
- // COPY (DestRC). If this condition fails, folding would be illegal.
- const MCInstrDesc &MovDesc = TII->get(MovOp);
- assert(MovDesc.getNumDefs() > 0 && MovDesc.operands()[0].RegClass != -1);
- const TargetRegisterClass *ResRC =
- TRI->getRegClass(MovDesc.operands()[0].RegClass);
- if (!DestRC->hasSuperClassEq(ResRC))
- return;
+ if (UseSubReg)
+ MovSrcRC = TRI->getMatchingSuperRegClass(SrcRC, MovSrcRC, UseSubReg);
+ if (!MRI->constrainRegClass(SrcReg, MovSrcRC))
+ break;
- MachineInstr::mop_iterator ImpOpI = UseMI->implicit_operands().begin();
- MachineInstr::mop_iterator ImpOpE = UseMI->implicit_operands().end();
- while (ImpOpI != ImpOpE) {
- MachineInstr::mop_iterator Tmp = ImpOpI;
- ImpOpI++;
- UseMI->removeOperand(UseMI->getOperandNo(Tmp));
- }
- UseMI->setDesc(TII->get(MovOp));
-
- if (MovOp == AMDGPU::V_MOV_B16_t16_e64) {
- const auto &SrcOp = UseMI->getOperand(UseOpIdx);
- MachineOperand NewSrcOp(SrcOp);
- MachineFunction *MF = UseMI->getParent()->getParent();
- UseMI->removeOperand(1);
- UseMI->addOperand(*MF, MachineOperand::CreateImm(0)); // src0_modifiers
- UseMI->addOperand(NewSrcOp); // src0
- UseMI->addOperand(*MF, MachineOperand::CreateImm(0)); // op_sel
- UseOpIdx = 2;
- UseOp = &UseMI->getOperand(UseOpIdx);
+ MachineInstr::mop_iterator ImpOpI = UseMI->implicit_operands().begin();
+ MachineInstr::mop_iterator ImpOpE = UseMI->implicit_operands().end();
+ while (ImpOpI != ImpOpE) {
+ MachineInstr::mop_iterator Tmp = ImpOpI;
+ ImpOpI++;
+ UseMI->removeOperand(UseMI->getOperandNo(Tmp));
+ }
+ UseMI->setDesc(MovDesc);
+
+ if (MovOp == AMDGPU::V_MOV_B16_t16_e64) {
+ const auto &SrcOp = UseMI->getOperand(UseOpIdx);
+ MachineOperand NewSrcOp(SrcOp);
+ MachineFunction *MF = UseMI->getParent()->getParent();
+ UseMI->removeOperand(1);
+ UseMI->addOperand(*MF, MachineOperand::CreateImm(0)); // src0_modifiers
+ UseMI->addOperand(NewSrcOp); // src0
+ UseMI->addOperand(*MF, MachineOperand::CreateImm(0)); // op_sel
+ UseOpIdx = SrcIdx;
+ UseOp = &UseMI->getOperand(UseOpIdx);
+ }
+ CopiesToReplace.push_back(UseMI);
+ break;
}
- CopiesToReplace.push_back(UseMI);
+
+ // We failed to replace the copy, so give up.
+ if (UseMI->getOpcode() == AMDGPU::COPY)
+ return;
+
} else {
if (UseMI->isCopy() && OpToFold.isReg() &&
UseMI->getOperand(0).getReg().isVirtual() &&
diff --git a/llvm/test/CodeGen/AMDGPU/br_cc.f16.ll b/llvm/test/CodeGen/AMDGPU/br_cc.f16.ll
index 2761cba5ea71b..bfef88cdba9ed 100644
--- a/llvm/test/CodeGen/AMDGPU/br_cc.f16.ll
+++ b/llvm/test/CodeGen/AMDGPU/br_cc.f16.ll
@@ -197,7 +197,7 @@ define amdgpu_kernel void @br_cc_f16_imm_a(
; GFX11-TRUE16-NEXT: v_cmp_nlt_f16_e32 vcc_lo, 0.5, v1.l
; GFX11-TRUE16-NEXT: s_cbranch_vccnz .LBB1_2
; GFX11-TRUE16-NEXT: ; %bb.1: ; %one
-; GFX11-TRUE16-NEXT: v_mov_b32_e32 v0, 0x3800
+; GFX11-TRUE16-NEXT: v_mov_b16_e32 v0.l, 0x3800
; GFX11-TRUE16-NEXT: .LBB1_2: ; %two
; GFX11-TRUE16-NEXT: s_mov_b32 s2, s6
; GFX11-TRUE16-NEXT: s_mov_b32 s3, s7
@@ -303,7 +303,7 @@ define amdgpu_kernel void @br_cc_f16_imm_b(
; GFX11-TRUE16-NEXT: v_cmp_ngt_f16_e32 vcc_lo, 0.5, v1.l
; GFX11-TRUE16-NEXT: s_cbranch_vccz .LBB2_2
; GFX11-TRUE16-NEXT: ; %bb.1: ; %two
-; GFX11-TRUE16-NEXT: v_mov_b32_e32 v0, 0x3800
+; GFX11-TRUE16-NEXT: v_mov_b16_e32 v0.l, 0x3800
; GFX11-TRUE16-NEXT: .LBB2_2: ; %one
; GFX11-TRUE16-NEXT: s_mov_b32 s2, s6
; GFX11-TRUE16-NEXT: s_mov_b32 s3, s7
diff --git a/llvm/test/CodeGen/AMDGPU/call-argument-types.ll b/llvm/test/CodeGen/AMDGPU/call-argument-types.ll
index 2a1be99dff5d2..b8dd377377dab 100644
--- a/llvm/test/CodeGen/AMDGPU/call-argument-types.ll
+++ b/llvm/test/CodeGen/AMDGPU/call-argument-types.ll
@@ -426,16 +426,27 @@ define amdgpu_kernel void @test_call_external_void_func_i8_imm(i32) #0 {
; GFX9-NEXT: s_swappc_b64 s[30:31], s[4:5]
; GFX9-NEXT: s_endpgm
;
-; GFX11-LABEL: test_call_external_void_func_i8_imm:
-; GFX11: ; %bb.0:
-; GFX11-NEXT: v_mov_b32_e32 v0, 0x7b
-; GFX11-NEXT: s_getpc_b64 s[2:3]
-; GFX11-NEXT: s_add_u32 s2, s2, external_void_func_i8 at rel32@lo+4
-; GFX11-NEXT: s_addc_u32 s3, s3, external_void_func_i8 at rel32@hi+12
-; GFX11-NEXT: s_mov_b64 s[6:7], s[0:1]
-; GFX11-NEXT: s_mov_b32 s32, 0
-; GFX11-NEXT: s_swappc_b64 s[30:31], s[2:3]
-; GFX11-NEXT: s_endpgm
+; GFX11-TRUE16-LABEL: test_call_external_void_func_i8_imm:
+; GFX11-TRUE16: ; %bb.0:
+; GFX11-TRUE16-NEXT: v_mov_b16_e32 v0.l, 0x7b
+; GFX11-TRUE16-NEXT: s_getpc_b64 s[2:3]
+; GFX11-TRUE16-NEXT: s_add_u32 s2, s2, external_void_func_i8 at rel32@lo+4
+; GFX11-TRUE16-NEXT: s_addc_u32 s3, s3, external_void_func_i8 at rel32@hi+12
+; GFX11-TRUE16-NEXT: s_mov_b64 s[6:7], s[0:1]
+; GFX11-TRUE16-NEXT: s_mov_b32 s32, 0
+; GFX11-TRUE16-NEXT: s_swappc_b64 s[30:31], s[2:3]
+; GFX11-TRUE16-NEXT: s_endpgm
+;
+; GFX11-FAKE16-LABEL: test_call_external_void_func_i8_imm:
+; GFX11-FAKE16: ; %bb.0:
+; GFX11-FAKE16-NEXT: v_mov_b32_e32 v0, 0x7b
+; GFX11-FAKE16-NEXT: s_getpc_b64 s[2:3]
+; GFX11-FAKE16-NEXT: s_add_u32 s2, s2, external_void_func_i8 at rel32@lo+4
+; GFX11-FAKE16-NEXT: s_addc_u32 s3, s3, external_void_func_i8 at rel32@hi+12
+; GFX11-FAKE16-NEXT: s_mov_b64 s[6:7], s[0:1]
+; GFX11-FAKE16-NEXT: s_mov_b32 s32, 0
+; GFX11-FAKE16-NEXT: s_swappc_b64 s[30:31], s[2:3]
+; GFX11-FAKE16-NEXT: s_endpgm
;
; HSA-LABEL: test_call_external_void_func_i8_imm:
; HSA: ; %bb.0:
@@ -723,16 +734,27 @@ define amdgpu_kernel void @test_call_external_void_func_i16_imm() #0 {
; GFX9-NEXT: s_swappc_b64 s[30:31], s[4:5]
; GFX9-NEXT: s_endpgm
;
-; GFX11-LABEL: test_call_external_void_func_i16_imm:
-; GFX11: ; %bb.0:
-; GFX11-NEXT: v_mov_b32_e32 v0, 0x7b
-; GFX11-NEXT: s_getpc_b64 s[2:3]
-; GFX11-NEXT: s_add_u32 s2, s2, external_void_func_i16 at rel32@lo+4
-; GFX11-NEXT: s_addc_u32 s3, s3, external_void_func_i16 at rel32@hi+12
-; GFX11-NEXT: s_mov_b64 s[6:7], s[0:1]
-; GFX11-NEXT: s_mov_b32 s32, 0
-; GFX11-NEXT: s_swappc_b64 s[30:31], s[2:3]
-; GFX11-NEXT: s_endpgm
+; GFX11-TRUE16-LABEL: test_call_external_void_func_i16_imm:
+; GFX11-TRUE16: ; %bb.0:
+; GFX11-TRUE16-NEXT: v_mov_b16_e32 v0.l, 0x7b
+; GFX11-TRUE16-NEXT: s_getpc_b64 s[2:3]
+; GFX11-TRUE16-NEXT: s_add_u32 s2, s2, external_void_func_i16 at rel32@lo+4
+; GFX11-TRUE16-NEXT: s_addc_u32 s3, s3, external_void_func_i16 at rel32@hi+12
+; GFX11-TRUE16-NEXT: s_mov_b64 s[6:7], s[0:1]
+; GFX11-TRUE16-NEXT: s_mov_b32 s32, 0
+; GFX11-TRUE16-NEXT: s_swappc_b64 s[30:31], s[2:3]
+; GFX11-TRUE16-NEXT: s_endpgm
+;
+; GFX11-FAKE16-LABEL: test_call_external_void_func_i16_imm:
+; GFX11-FAKE16: ; %bb.0:
+; GFX11-FAKE16-NEXT: v_mov_b32_e32 v0, 0x7b
+; GFX11-FAKE16-NEXT: s_getpc_b64 s[2:3]
+; GFX11-FAKE16-NEXT: s_add_u32 s2, s2, external_void_func_i16 at rel32@lo+4
+; GFX11-FAKE16-NEXT: s_addc_u32 s3, s3, external_void_func_i16 at rel32@hi+12
+; GFX11-FAKE16-NEXT: s_mov_b64 s[6:7], s[0:1]
+; GFX11-FAKE16-NEXT: s_mov_b32 s32, 0
+; GFX11-FAKE16-NEXT: s_swappc_b64 s[30:31], s[2:3]
+; GFX11-FAKE16-NEXT: s_endpgm
;
; HSA-LABEL: test_call_external_void_func_i16_imm:
; HSA: ; %bb.0:
@@ -1642,16 +1664,27 @@ define amdgpu_kernel void @test_call_external_void_func_f16_imm() #0 {
; GFX9-NEXT: s_swappc_b64 s[30:31], s[4:5]
; GFX9-NEXT: s_endpgm
;
-; GFX11-LABEL: test_call_external_void_func_f16_imm:
-; GFX11: ; %bb.0:
-; GFX11-NEXT: v_mov_b32_e32 v0, 0x4400
-; GFX11-NEXT: s_getpc_b64 s[2:3]
-; GFX11-NEXT: s_add_u32 s2, s2, external_void_func_f16 at rel32@lo+4
-; GFX11-NEXT: s_addc_u32 s3, s3, external_void_func_f16 at rel32@hi+12
-; GFX11-NEXT: s_mov_b64 s[6:7], s[0:1]
-; GFX11-NEXT: s_mov_b32 s32, 0
-; GFX11-NEXT: s_swappc_b64 s[30:31], s[2:3]
-; GFX11-NEXT: s_endpgm
+; GFX11-TRUE16-LABEL: test_call_external_void_func_f16_imm:
+; GFX11-TRUE16: ; %bb.0:
+; GFX11-TRUE16-NEXT: v_mov_b16_e32 v0.l, 0x4400
+; GFX11-TRUE16-NEXT: s_getpc_b64 s[2:3]
+; GFX11-TRUE16-NEXT: s_add_u32 s2, s2, external_void_func_f16 at rel32@lo+4
+; GFX11-TRUE16-NEXT: s_addc_u32 s3, s3, external_void_func_f16 at rel32@hi+12
+; GFX11-TRUE16-NEXT: s_mov_b64 s[6:7], s[0:1]
+; GFX11-TRUE16-NEXT: s_mov_b32 s32, 0
+; GFX11-TRUE16-NEXT: s_swappc_b64 s[30:31], s[2:3]
+; GFX11-TRUE16-NEXT: s_endpgm
+;
+; GFX11-FAKE16-LABEL: test_call_external_void_func_f16_imm:
+; GFX11-FAKE16: ; %bb.0:
+; GFX11-FAKE16-NEXT: v_mov_b32_e32 v0, 0x4400
+; GFX11-FAKE16-NEXT: s_getpc_b64 s[2:3]
+; GFX11-FAKE16-NEXT: s_add_u32 s2, s2, external_void_func_f16 at rel32@lo+4
+; GFX11-FAKE16-NEXT: s_addc_u32 s3, s3, external_void_func_f16 at rel32@hi+12
+; GFX11-FAKE16-NEXT: s_mov_b64 s[6:7], s[0:1]
+; GFX11-FAKE16-NEXT: s_mov_b32 s32, 0
+; GFX11-FAKE16-NEXT: s_swappc_b64 s[30:31], s[2:3]
+; GFX11-FAKE16-NEXT: s_endpgm
;
; HSA-LABEL: test_call_external_void_func_f16_imm:
; HSA: ; %bb.0:
diff --git a/llvm/test/CodeGen/AMDGPU/gfx-callable-argument-types.ll b/llvm/test/CodeGen/AMDGPU/gfx-callable-argument-types.ll
index 919464a936740..f67ab18dd8ef1 100644
--- a/llvm/test/CodeGen/AMDGPU/gfx-callable-argument-types.ll
+++ b/llvm/test/CodeGen/AMDGPU/gfx-callable-argument-types.ll
@@ -559,33 +559,61 @@ define amdgpu_gfx void @test_call_external_void_func_i8_imm(i32) #0 {
; GFX10-NEXT: s_waitcnt vmcnt(0)
; GFX10-NEXT: s_setpc_b64 s[30:31]
;
-; GFX11-LABEL: test_call_external_void_func_i8_imm:
-; GFX11: ; %bb.0:
-; GFX11-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0)
-; GFX11-NEXT: s_mov_b32 s0, s33
-; GFX11-NEXT: s_mov_b32 s33, s32
-; GFX11-NEXT: s_or_saveexec_b32 s1, -1
-; GFX11-NEXT: scratch_store_b32 off, v40, s33 ; 4-byte Folded Spill
-; GFX11-NEXT: s_mov_b32 exec_lo, s1
-; GFX11-NEXT: v_writelane_b32 v40, s0, 2
-; GFX11-NEXT: v_mov_b32_e32 v0, 0x7b
-; GFX11-NEXT: s_mov_b32 s1, external_void_func_i8 at abs32@hi
-; GFX11-NEXT: s_mov_b32 s0, external_void_func_i8 at abs32@lo
-; GFX11-NEXT: s_add_i32 s32, s32, 16
-; GFX11-NEXT: v_writelane_b32 v40, s30, 0
-; GFX11-NEXT: v_writelane_b32 v40, s31, 1
-; GFX11-NEXT: s_swappc_b64 s[30:31], s[0:1]
-; GFX11-NEXT: s_delay_alu instid0(VALU_DEP_1)
-; GFX11-NEXT: v_readlane_b32 s31, v40, 1
-; GFX11-NEXT: v_readlane_b32 s30, v40, 0
-; GFX11-NEXT: s_mov_b32 s32, s33
-; GFX11-NEXT: v_readlane_b32 s0, v40, 2
-; GFX11-NEXT: s_or_saveexec_b32 s1, -1
-; GFX11-NEXT: scratch_load_b32 v40, off, s33 ; 4-byte Folded Reload
-; GFX11-NEXT: s_mov_b32 exec_lo, s1
-; GFX11-NEXT: s_mov_b32 s33, s0
-; GFX11-NEXT: s_waitcnt vmcnt(0)
-; GFX11-NEXT: s_setpc_b64 s[30:31]
+; GFX11-TRUE16-LABEL: test_call_external_void_func_i8_imm:
+; GFX11-TRUE16: ; %bb.0:
+; GFX11-TRUE16-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0)
+; GFX11-TRUE16-NEXT: s_mov_b32 s0, s33
+; GFX11-TRUE16-NEXT: s_mov_b32 s33, s32
+; GFX11-TRUE16-NEXT: s_or_saveexec_b32 s1, -1
+; GFX11-TRUE16-NEXT: scratch_store_b32 off, v40, s33 ; 4-byte Folded Spill
+; GFX11-TRUE16-NEXT: s_mov_b32 exec_lo, s1
+; GFX11-TRUE16-NEXT: v_writelane_b32 v40, s0, 2
+; GFX11-TRUE16-NEXT: v_mov_b16_e32 v0.l, 0x7b
+; GFX11-TRUE16-NEXT: s_mov_b32 s1, external_void_func_i8 at abs32@hi
+; GFX11-TRUE16-NEXT: s_mov_b32 s0, external_void_func_i8 at abs32@lo
+; GFX11-TRUE16-NEXT: s_add_i32 s32, s32, 16
+; GFX11-TRUE16-NEXT: v_writelane_b32 v40, s30, 0
+; GFX11-TRUE16-NEXT: v_writelane_b32 v40, s31, 1
+; GFX11-TRUE16-NEXT: s_swappc_b64 s[30:31], s[0:1]
+; GFX11-TRUE16-NEXT: s_delay_alu instid0(VALU_DEP_1)
+; GFX11-TRUE16-NEXT: v_readlane_b32 s31, v40, 1
+; GFX11-TRUE16-NEXT: v_readlane_b32 s30, v40, 0
+; GFX11-TRUE16-NEXT: s_mov_b32 s32, s33
+; GFX11-TRUE16-NEXT: v_readlane_b32 s0, v40, 2
+; GFX11-TRUE16-NEXT: s_or_saveexec_b32 s1, -1
+; GFX11-TRUE16-NEXT: scratch_load_b32 v40, off, s33 ; 4-byte Folded Reload
+; GFX11-TRUE16-NEXT: s_mov_b32 exec_lo, s1
+; GFX11-TRUE16-NEXT: s_mov_b32 s33, s0
+; GFX11-TRUE16-NEXT: s_waitcnt vmcnt(0)
+; GFX11-TRUE16-NEXT: s_setpc_b64 s[30:31]
+;
+; GFX11-FAKE16-LABEL: test_call_external_void_func_i8_imm:
+; GFX11-FAKE16: ; %bb.0:
+; GFX11-FAKE16-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0)
+; GFX11-FAKE16-NEXT: s_mov_b32 s0, s33
+; GFX11-FAKE16-NEXT: s_mov_b32 s33, s32
+; GFX11-FAKE16-NEXT: s_or_saveexec_b32 s1, -1
+; GFX11-FAKE16-NEXT: scratch_store_b32 off, v40, s33 ; 4-byte Folded Spill
+; GFX11-FAKE16-NEXT: s_mov_b32 exec_lo, s1
+; GFX11-FAKE16-NEXT: v_writelane_b32 v40, s0, 2
+; GFX11-FAKE16-NEXT: v_mov_b32_e32 v0, 0x7b
+; GFX11-FAKE16-NEXT: s_mov_b32 s1, external_void_func_i8 at abs32@hi
+; GFX11-FAKE16-NEXT: s_mov_b32 s0, external_void_func_i8 at abs32@lo
+; GFX11-FAKE16-NEXT: s_add_i32 s32, s32, 16
+; GFX11-FAKE16-NEXT: v_writelane_b32 v40, s30, 0
+; GFX11-FAKE16-NEXT: v_writelane_b32 v40, s31, 1
+; GFX11-FAKE16-NEXT: s_swappc_b64 s[30:31], s[0:1]
+; GFX11-FAKE16-NEXT: s_delay_alu instid0(VALU_DEP_1)
+; GFX11-FAKE16-NEXT: v_readlane_b32 s31, v40, 1
+; GFX11-FAKE16-NEXT: v_readlane_b32 s30, v40, 0
+; GFX11-FAKE16-NEXT: s_mov_b32 s32, s33
+; GFX11-FAKE16-NEXT: v_readlane_b32 s0, v40, 2
+; GFX11-FAKE16-NEXT: s_or_saveexec_b32 s1, -1
+; GFX11-FAKE16-NEXT: scratch_load_b32 v40, off, s33 ; 4-byte Folded Reload
+; GFX11-FAKE16-NEXT: s_mov_b32 exec_lo, s1
+; GFX11-FAKE16-NEXT: s_mov_b32 s33, s0
+; GFX11-FAKE16-NEXT: s_waitcnt vmcnt(0)
+; GFX11-FAKE16-NEXT: s_setpc_b64 s[30:31]
;
; GFX10-SCRATCH-LABEL: test_call_external_void_func_i8_imm:
; GFX10-SCRATCH: ; %bb.0:
@@ -978,33 +1006,61 @@ define amdgpu_gfx void @test_call_external_void_func_i16_imm() #0 {
; GFX10-NEXT: s_waitcnt vmcnt(0)
; GFX10-NEXT: s_setpc_b64 s[30:31]
;
-; GFX11-LABEL: test_call_external_void_func_i16_imm:
-; GFX11: ; %bb.0:
-; GFX11-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0)
-; GFX11-NEXT: s_mov_b32 s0, s33
-; GFX11-NEXT: s_mov_b32 s33, s32
-; GFX11-NEXT: s_or_saveexec_b32 s1, -1
-; GFX11-NEXT: scratch_store_b32 off, v40, s33 ; 4-byte Folded Spill
-; GFX11-NEXT: s_mov_b32 exec_lo, s1
-; GFX11-NEXT: v_writelane_b32 v40, s0, 2
-; GFX11-NEXT: v_mov_b32_e32 v0, 0x7b
-; GFX11-NEXT: s_mov_b32 s1, external_void_func_i16 at abs32@hi
-; GFX11-NEXT: s_mov_b32 s0, external_void_func_i16 at abs32@lo
-; GFX11-NEXT: s_add_i32 s32, s32, 16
-; GFX11-NEXT: v_writelane_b32 v40, s30, 0
-; GFX11-NEXT: v_writelane_b32 v40, s31, 1
-; GFX11-NEXT: s_swappc_b64 s[30:31], s[0:1]
-; GFX11-NEXT: s_delay_alu instid0(VALU_DEP_1)
-; GFX11-NEXT: v_readlane_b32 s31, v40, 1
-; GFX11-NEXT: v_readlane_b32 s30, v40, 0
-; GFX11-NEXT: s_mov_b32 s32, s33
-; GFX11-NEXT: v_readlane_b32 s0, v40, 2
-; GFX11-NEXT: s_or_saveexec_b32 s1, -1
-; GFX11-NEXT: scratch_load_b32 v40, off, s33 ; 4-byte Folded Reload
-; GFX11-NEXT: s_mov_b32 exec_lo, s1
-; GFX11-NEXT: s_mov_b32 s33, s0
-; GFX11-NEXT: s_waitcnt vmcnt(0)
-; GFX11-NEXT: s_setpc_b64 s[30:31]
+; GFX11-TRUE16-LABEL: test_call_external_void_func_i16_imm:
+; GFX11-TRUE16: ; %bb.0:
+; GFX11-TRUE16-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0)
+; GFX11-TRUE16-NEXT: s_mov_b32 s0, s33
+; GFX11-TRUE16-NEXT: s_mov_b32 s33, s32
+; GFX11-TRUE16-NEXT: s_or_saveexec_b32 s1, -1
+; GFX11-TRUE16-NEXT: scratch_store_b32 off, v40, s33 ; 4-byte Folded Spill
+; GFX11-TRUE16-NEXT: s_mov_b32 exec_lo, s1
+; GFX11-TRUE16-NEXT: v_writelane_b32 v40, s0, 2
+; GFX11-TRUE16-NEXT: v_mov_b16_e32 v0.l, 0x7b
+; GFX11-TRUE16-NEXT: s_mov_b32 s1, external_void_func_i16 at abs32@hi
+; GFX11-TRUE16-NEXT: s_mov_b32 s0, external_void_func_i16 at abs32@lo
+; GFX11-TRUE16-NEXT: s_add_i32 s32, s32, 16
+; GFX11-TRUE16-NEXT: v_writelane_b32 v40, s30, 0
+; GFX11-TRUE16-NEXT: v_writelane_b32 v40, s31, 1
+; GFX11-TRUE16-NEXT: s_swappc_b64 s[30:31], s[0:1]
+; GFX11-TRUE16-NEXT: s_delay_alu instid0(VALU_DEP_1)
+; GFX11-TRUE16-NEXT: v_readlane_b32 s31, v40, 1
+; GFX11-TRUE16-NEXT: v_readlane_b32 s30, v40, 0
+; GFX11-TRUE16-NEXT: s_mov_b32 s32, s33
+; GFX11-TRUE16-NEXT: v_readlane_b32 s0, v40, 2
+; GFX11-TRUE16-NEXT: s_or_saveexec_b32 s1, -1
+; GFX11-TRUE16-NEXT: scratch_load_b32 v40, off, s33 ; 4-byte Folded Reload
+; GFX11-TRUE16-NEXT: s_mov_b32 exec_lo, s1
+; GFX11-TRUE16-NEXT: s_mov_b32 s33, s0
+; GFX11-TRUE16-NEXT: s_waitcnt vmcnt(0)
+; GFX11-TRUE16-NEXT: s_setpc_b64 s[30:31]
+;
+; GFX11-FAKE16-LABEL: test_call_external_void_func_i16_imm:
+; GFX11-FAKE16: ; %bb.0:
+; GFX11-FAKE16-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0)
+; GFX11-FAKE16-NEXT: s_mov_b32 s0, s33
+; GFX11-FAKE16-NEXT: s_mov_b32 s33, s32
+; GFX11-FAKE16-NEXT: s_or_saveexec_b32 s1, -1
+; GFX11-FAKE16-NEXT: scratch_store_b32 off, v40, s33 ; 4-byte Folded Spill
+; GFX11-FAKE16-NEXT: s_mov_b32 exec_lo, s1
+; GFX11-FAKE16-NEXT: v_writelane_b32 v40, s0, 2
+; GFX11-FAKE16-NEXT: v_mov_b32_e32 v0, 0x7b
+; GFX11-FAKE16-NEXT: s_mov_b32 s1, external_void_func_i16 at abs32@hi
+; GFX11-FAKE16-NEXT: s_mov_b32 s0, external_void_func_i16 at abs32@lo
+; GFX11-FAKE16-NEXT: s_add_i32 s32, s32, 16
+; GFX11-FAKE16-NEXT: v_writelane_b32 v40, s30, 0
+; GFX11-FAKE16-NEXT: v_writelane_b32 v40, s31, 1
+; GFX11-FAKE16-NEXT: s_swappc_b64 s[30:31], s[0:1]
+; GFX11-FAKE16-NEXT: s_delay_alu instid0(VALU_DEP_1)
+; GFX11-FAKE16-NEXT: v_readlane_b32 s31, v40, 1
+; GFX11-FAKE16-NEXT: v_readlane_b32 s30, v40, 0
+; GFX11-FAKE16-NEXT: s_mov_b32 s32, s33
+; GFX11-FAKE16-NEXT: v_readlane_b32 s0, v40, 2
+; GFX11-FAKE16-NEXT: s_or_saveexec_b32 s1, -1
+; GFX11-FAKE16-NEXT: scratch_load_b32 v40, off, s33 ; 4-byte Folded Reload
+; GFX11-FAKE16-NEXT: s_mov_b32 exec_lo, s1
+; GFX11-FAKE16-NEXT: s_mov_b32 s33, s0
+; GFX11-FAKE16-NEXT: s_waitcnt vmcnt(0)
+; GFX11-FAKE16-NEXT: s_setpc_b64 s[30:31]
;
; GFX10-SCRATCH-LABEL: test_call_external_void_func_i16_imm:
; GFX10-SCRATCH: ; %bb.0:
@@ -2161,33 +2217,61 @@ define amdgpu_gfx void @test_call_external_void_func_f16_imm() #0 {
; GFX10-NEXT: s_waitcnt vmcnt(0)
; GFX10-NEXT: s_setpc_b64 s[30:31]
;
-; GFX11-LABEL: test_call_external_void_func_f16_imm:
-; GFX11: ; %bb.0:
-; GFX11-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0)
-; GFX11-NEXT: s_mov_b32 s0, s33
-; GFX11-NEXT: s_mov_b32 s33, s32
-; GFX11-NEXT: s_or_saveexec_b32 s1, -1
-; GFX11-NEXT: scratch_store_b32 off, v40, s33 ; 4-byte Folded Spill
-; GFX11-NEXT: s_mov_b32 exec_lo, s1
-; GFX11-NEXT: v_writelane_b32 v40, s0, 2
-; GFX11-NEXT: v_mov_b32_e32 v0, 0x4400
-; GFX11-NEXT: s_mov_b32 s1, external_void_func_f16 at abs32@hi
-; GFX11-NEXT: s_mov_b32 s0, external_void_func_f16 at abs32@lo
-; GFX11-NEXT: s_add_i32 s32, s32, 16
-; GFX11-NEXT: v_writelane_b32 v40, s30, 0
-; GFX11-NEXT: v_writelane_b32 v40, s31, 1
-; GFX11-NEXT: s_swappc_b64 s[30:31], s[0:1]
-; GFX11-NEXT: s_delay_alu instid0(VALU_DEP_1)
-; GFX11-NEXT: v_readlane_b32 s31, v40, 1
-; GFX11-NEXT: v_readlane_b32 s30, v40, 0
-; GFX11-NEXT: s_mov_b32 s32, s33
-; GFX11-NEXT: v_readlane_b32 s0, v40, 2
-; GFX11-NEXT: s_or_saveexec_b32 s1, -1
-; GFX11-NEXT: scratch_load_b32 v40, off, s33 ; 4-byte Folded Reload
-; GFX11-NEXT: s_mov_b32 exec_lo, s1
-; GFX11-NEXT: s_mov_b32 s33, s0
-; GFX11-NEXT: s_waitcnt vmcnt(0)
-; GFX11-NEXT: s_setpc_b64 s[30:31]
+; GFX11-TRUE16-LABEL: test_call_external_void_func_f16_imm:
+; GFX11-TRUE16: ; %bb.0:
+; GFX11-TRUE16-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0)
+; GFX11-TRUE16-NEXT: s_mov_b32 s0, s33
+; GFX11-TRUE16-NEXT: s_mov_b32 s33, s32
+; GFX11-TRUE16-NEXT: s_or_saveexec_b32 s1, -1
+; GFX11-TRUE16-NEXT: scratch_store_b32 off, v40, s33 ; 4-byte Folded Spill
+; GFX11-TRUE16-NEXT: s_mov_b32 exec_lo, s1
+; GFX11-TRUE16-NEXT: v_writelane_b32 v40, s0, 2
+; GFX11-TRUE16-NEXT: v_mov_b16_e32 v0.l, 0x4400
+; GFX11-TRUE16-NEXT: s_mov_b32 s1, external_void_func_f16 at abs32@hi
+; GFX11-TRUE16-NEXT: s_mov_b32 s0, external_void_func_f16 at abs32@lo
+; GFX11-TRUE16-NEXT: s_add_i32 s32, s32, 16
+; GFX11-TRUE16-NEXT: v_writelane_b32 v40, s30, 0
+; GFX11-TRUE16-NEXT: v_writelane_b32 v40, s31, 1
+; GFX11-TRUE16-NEXT: s_swappc_b64 s[30:31], s[0:1]
+; GFX11-TRUE16-NEXT: s_delay_alu instid0(VALU_DEP_1)
+; GFX11-TRUE16-NEXT: v_readlane_b32 s31, v40, 1
+; GFX11-TRUE16-NEXT: v_readlane_b32 s30, v40, 0
+; GFX11-TRUE16-NEXT: s_mov_b32 s32, s33
+; GFX11-TRUE16-NEXT: v_readlane_b32 s0, v40, 2
+; GFX11-TRUE16-NEXT: s_or_saveexec_b32 s1, -1
+; GFX11-TRUE16-NEXT: scratch_load_b32 v40, off, s33 ; 4-byte Folded Reload
+; GFX11-TRUE16-NEXT: s_mov_b32 exec_lo, s1
+; GFX11-TRUE16-NEXT: s_mov_b32 s33, s0
+; GFX11-TRUE16-NEXT: s_waitcnt vmcnt(0)
+; GFX11-TRUE16-NEXT: s_setpc_b64 s[30:31]
+;
+; GFX11-FAKE16-LABEL: test_call_external_void_func_f16_imm:
+; GFX11-FAKE16: ; %bb.0:
+; GFX11-FAKE16-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0)
+; GFX11-FAKE16-NEXT: s_mov_b32 s0, s33
+; GFX11-FAKE16-NEXT: s_mov_b32 s33, s32
+; GFX11-FAKE16-NEXT: s_or_saveexec_b32 s1, -1
+; GFX11-FAKE16-NEXT: scratch_store_b32 off, v40, s33 ; 4-byte Folded Spill
+; GFX11-FAKE16-NEXT: s_mov_b32 exec_lo, s1
+; GFX11-FAKE16-NEXT: v_writelane_b32 v40, s0, 2
+; GFX11-FAKE16-NEXT: v_mov_b32_e32 v0, 0x4400
+; GFX11-FAKE16-NEXT: s_mov_b32 s1, external_void_func_f16 at abs32@hi
+; GFX11-FAKE16-NEXT: s_mov_b32 s0, external_void_func_f16 at abs32@lo
+; GFX11-FAKE16-NEXT: s_add_i32 s32, s32, 16
+; GFX11-FAKE16-NEXT: v_writelane_b32 v40, s30, 0
+; GFX11-FAKE16-NEXT: v_writelane_b32 v40, s31, 1
+; GFX11-FAKE16-NEXT: s_swappc_b64 s[30:31], s[0:1]
+; GFX11-FAKE16-NEXT: s_delay_alu instid0(VALU_DEP_1)
+; GFX11-FAKE16-NEXT: v_readlane_b32 s31, v40, 1
+; GFX11-FAKE16-NEXT: v_readlane_b32 s30, v40, 0
+; GFX11-FAKE16-NEXT: s_mov_b32 s32, s33
+; GFX11-FAKE16-NEXT: v_readlane_b32 s0, v40, 2
+; GFX11-FAKE16-NEXT: s_or_saveexec_b32 s1, -1
+; GFX11-FAKE16-NEXT: scratch_load_b32 v40, off, s33 ; 4-byte Folded Reload
+; GFX11-FAKE16-NEXT: s_mov_b32 exec_lo, s1
+; GFX11-FAKE16-NEXT: s_mov_b32 s33, s0
+; GFX11-FAKE16-NEXT: s_waitcnt vmcnt(0)
+; GFX11-FAKE16-NEXT: s_setpc_b64 s[30:31]
;
; GFX10-SCRATCH-LABEL: test_call_external_void_func_f16_imm:
; GFX10-SCRATCH: ; %bb.0:
diff --git a/llvm/test/CodeGen/AMDGPU/gfx-callable-return-types.ll b/llvm/test/CodeGen/AMDGPU/gfx-callable-return-types.ll
index 5c183f5f2ebe6..b750d28ffa7d3 100644
--- a/llvm/test/CodeGen/AMDGPU/gfx-callable-return-types.ll
+++ b/llvm/test/CodeGen/AMDGPU/gfx-callable-return-types.ll
@@ -106,11 +106,17 @@ define amdgpu_gfx i16 @return_i16() #0 {
; GFX9-NEXT: v_mov_b32_e32 v0, 10
; GFX9-NEXT: s_setpc_b64 s[30:31]
;
-; GFX10PLUS-LABEL: return_i16:
-; GFX10PLUS: ; %bb.0: ; %entry
-; GFX10PLUS-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0)
-; GFX10PLUS-NEXT: v_mov_b32_e32 v0, 10
-; GFX10PLUS-NEXT: s_setpc_b64 s[30:31]
+; GFX10-LABEL: return_i16:
+; GFX10: ; %bb.0: ; %entry
+; GFX10-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0)
+; GFX10-NEXT: v_mov_b32_e32 v0, 10
+; GFX10-NEXT: s_setpc_b64 s[30:31]
+;
+; GFX11-LABEL: return_i16:
+; GFX11: ; %bb.0: ; %entry
+; GFX11-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0)
+; GFX11-NEXT: v_mov_b16_e32 v0.l, 10
+; GFX11-NEXT: s_setpc_b64 s[30:31]
entry:
ret i16 10
}
diff --git a/llvm/test/CodeGen/AMDGPU/imm16.ll b/llvm/test/CodeGen/AMDGPU/imm16.ll
index 676773a94d29a..91b9773ad159e 100644
--- a/llvm/test/CodeGen/AMDGPU/imm16.ll
+++ b/llvm/test/CodeGen/AMDGPU/imm16.ll
@@ -19,16 +19,27 @@ define amdgpu_kernel void @store_inline_imm_neg_0.0_i16(ptr addrspace(1) %out) {
; GFX10-NEXT: s_waitcnt_vscnt null, 0x0 ; encoding: [0x00,0x00,0xfd,0xbb]
; GFX10-NEXT: s_endpgm ; encoding: [0x00,0x00,0x81,0xbf]
;
-; GFX11-LABEL: store_inline_imm_neg_0.0_i16:
-; GFX11: ; %bb.0:
-; GFX11-NEXT: s_load_b64 s[0:1], s[4:5], 0x0 ; encoding: [0x02,0x00,0x04,0xf4,0x00,0x00,0x00,0xf8]
-; GFX11-NEXT: v_mov_b32_e32 v0, 0xffff8000 ; encoding: [0xff,0x02,0x00,0x7e,0x00,0x80,0xff,0xff]
-; GFX11-NEXT: s_mov_b32 s3, 0x31016000 ; encoding: [0xff,0x00,0x83,0xbe,0x00,0x60,0x01,0x31]
-; GFX11-NEXT: s_mov_b32 s2, -1 ; encoding: [0xc1,0x00,0x82,0xbe]
-; GFX11-NEXT: s_waitcnt lgkmcnt(0) ; encoding: [0x07,0xfc,0x89,0xbf]
-; GFX11-NEXT: buffer_store_b16 v0, off, s[0:3], 0 dlc ; encoding: [0x00,0x20,0x64,0xe0,0x00,0x00,0x00,0x80]
-; GFX11-NEXT: s_waitcnt_vscnt null, 0x0 ; encoding: [0x00,0x00,0x7c,0xbc]
-; GFX11-NEXT: s_endpgm ; encoding: [0x00,0x00,0xb0,0xbf]
+; GFX11-TRUE16-LABEL: store_inline_imm_neg_0.0_i16:
+; GFX11-TRUE16: ; %bb.0:
+; GFX11-TRUE16-NEXT: s_load_b64 s[0:1], s[4:5], 0x0 ; encoding: [0x02,0x00,0x04,0xf4,0x00,0x00,0x00,0xf8]
+; GFX11-TRUE16-NEXT: v_mov_b16_e32 v0.l, 0x8000 ; encoding: [0xff,0x38,0x00,0x7e,0x00,0x80,0xff,0xff]
+; GFX11-TRUE16-NEXT: s_mov_b32 s3, 0x31016000 ; encoding: [0xff,0x00,0x83,0xbe,0x00,0x60,0x01,0x31]
+; GFX11-TRUE16-NEXT: s_mov_b32 s2, -1 ; encoding: [0xc1,0x00,0x82,0xbe]
+; GFX11-TRUE16-NEXT: s_waitcnt lgkmcnt(0) ; encoding: [0x07,0xfc,0x89,0xbf]
+; GFX11-TRUE16-NEXT: buffer_store_b16 v0, off, s[0:3], 0 dlc ; encoding: [0x00,0x20,0x64,0xe0,0x00,0x00,0x00,0x80]
+; GFX11-TRUE16-NEXT: s_waitcnt_vscnt null, 0x0 ; encoding: [0x00,0x00,0x7c,0xbc]
+; GFX11-TRUE16-NEXT: s_endpgm ; encoding: [0x00,0x00,0xb0,0xbf]
+;
+; GFX11-FAKE16-LABEL: store_inline_imm_neg_0.0_i16:
+; GFX11-FAKE16: ; %bb.0:
+; GFX11-FAKE16-NEXT: s_load_b64 s[0:1], s[4:5], 0x0 ; encoding: [0x02,0x00,0x04,0xf4,0x00,0x00,0x00,0xf8]
+; GFX11-FAKE16-NEXT: v_mov_b32_e32 v0, 0xffff8000 ; encoding: [0xff,0x02,0x00,0x7e,0x00,0x80,0xff,0xff]
+; GFX11-FAKE16-NEXT: s_mov_b32 s3, 0x31016000 ; encoding: [0xff,0x00,0x83,0xbe,0x00,0x60,0x01,0x31]
+; GFX11-FAKE16-NEXT: s_mov_b32 s2, -1 ; encoding: [0xc1,0x00,0x82,0xbe]
+; GFX11-FAKE16-NEXT: s_waitcnt lgkmcnt(0) ; encoding: [0x07,0xfc,0x89,0xbf]
+; GFX11-FAKE16-NEXT: buffer_store_b16 v0, off, s[0:3], 0 dlc ; encoding: [0x00,0x20,0x64,0xe0,0x00,0x00,0x00,0x80]
+; GFX11-FAKE16-NEXT: s_waitcnt_vscnt null, 0x0 ; encoding: [0x00,0x00,0x7c,0xbc]
+; GFX11-FAKE16-NEXT: s_endpgm ; encoding: [0x00,0x00,0xb0,0xbf]
;
; VI-LABEL: store_inline_imm_neg_0.0_i16:
; VI: ; %bb.0:
@@ -66,15 +77,25 @@ define amdgpu_kernel void @store_inline_imm_0.0_f16(ptr addrspace(1) %out) {
; GFX10-NEXT: buffer_store_short v0, off, s[0:3], 0 ; encoding: [0x00,0x00,0x68,0xe0,0x00,0x00,0x00,0x80]
; GFX10-NEXT: s_endpgm ; encoding: [0x00,0x00,0x81,0xbf]
;
-; GFX11-LABEL: store_inline_imm_0.0_f16:
-; GFX11: ; %bb.0:
-; GFX11-NEXT: s_load_b64 s[0:1], s[4:5], 0x0 ; encoding: [0x02,0x00,0x04,0xf4,0x00,0x00,0x00,0xf8]
-; GFX11-NEXT: v_mov_b32_e32 v0, 0 ; encoding: [0x80,0x02,0x00,0x7e]
-; GFX11-NEXT: s_mov_b32 s3, 0x31016000 ; encoding: [0xff,0x00,0x83,0xbe,0x00,0x60,0x01,0x31]
-; GFX11-NEXT: s_mov_b32 s2, -1 ; encoding: [0xc1,0x00,0x82,0xbe]
-; GFX11-NEXT: s_waitcnt lgkmcnt(0) ; encoding: [0x07,0xfc,0x89,0xbf]
-; GFX11-NEXT: buffer_store_b16 v0, off, s[0:3], 0 ; encoding: [0x00,0x00,0x64,0xe0,0x00,0x00,0x00,0x80]
-; GFX11-NEXT: s_endpgm ; encoding: [0x00,0x00,0xb0,0xbf]
+; GFX11-TRUE16-LABEL: store_inline_imm_0.0_f16:
+; GFX11-TRUE16: ; %bb.0:
+; GFX11-TRUE16-NEXT: s_load_b64 s[0:1], s[4:5], 0x0 ; encoding: [0x02,0x00,0x04,0xf4,0x00,0x00,0x00,0xf8]
+; GFX11-TRUE16-NEXT: v_mov_b16_e32 v0.l, 0 ; encoding: [0x80,0x38,0x00,0x7e]
+; GFX11-TRUE16-NEXT: s_mov_b32 s3, 0x31016000 ; encoding: [0xff,0x00,0x83,0xbe,0x00,0x60,0x01,0x31]
+; GFX11-TRUE16-NEXT: s_mov_b32 s2, -1 ; encoding: [0xc1,0x00,0x82,0xbe]
+; GFX11-TRUE16-NEXT: s_waitcnt lgkmcnt(0) ; encoding: [0x07,0xfc,0x89,0xbf]
+; GFX11-TRUE16-NEXT: buffer_store_b16 v0, off, s[0:3], 0 ; encoding: [0x00,0x00,0x64,0xe0,0x00,0x00,0x00,0x80]
+; GFX11-TRUE16-NEXT: s_endpgm ; encoding: [0x00,0x00,0xb0,0xbf]
+;
+; GFX11-FAKE16-LABEL: store_inline_imm_0.0_f16:
+; GFX11-FAKE16: ; %bb.0:
+; GFX11-FAKE16-NEXT: s_load_b64 s[0:1], s[4:5], 0x0 ; encoding: [0x02,0x00,0x04,0xf4,0x00,0x00,0x00,0xf8]
+; GFX11-FAKE16-NEXT: v_mov_b32_e32 v0, 0 ; encoding: [0x80,0x02,0x00,0x7e]
+; GFX11-FAKE16-NEXT: s_mov_b32 s3, 0x31016000 ; encoding: [0xff,0x00,0x83,0xbe,0x00,0x60,0x01,0x31]
+; GFX11-FAKE16-NEXT: s_mov_b32 s2, -1 ; encoding: [0xc1,0x00,0x82,0xbe]
+; GFX11-FAKE16-NEXT: s_waitcnt lgkmcnt(0) ; encoding: [0x07,0xfc,0x89,0xbf]
+; GFX11-FAKE16-NEXT: buffer_store_b16 v0, off, s[0:3], 0 ; encoding: [0x00,0x00,0x64,0xe0,0x00,0x00,0x00,0x80]
+; GFX11-FAKE16-NEXT: s_endpgm ; encoding: [0x00,0x00,0xb0,0xbf]
;
; VI-LABEL: store_inline_imm_0.0_f16:
; VI: ; %bb.0:
@@ -110,15 +131,25 @@ define amdgpu_kernel void @store_imm_neg_0.0_f16(ptr addrspace(1) %out) {
; GFX10-NEXT: buffer_store_short v0, off, s[0:3], 0 ; encoding: [0x00,0x00,0x68,0xe0,0x00,0x00,0x00,0x80]
; GFX10-NEXT: s_endpgm ; encoding: [0x00,0x00,0x81,0xbf]
;
-; GFX11-LABEL: store_imm_neg_0.0_f16:
-; GFX11: ; %bb.0:
-; GFX11-NEXT: s_load_b64 s[0:1], s[4:5], 0x0 ; encoding: [0x02,0x00,0x04,0xf4,0x00,0x00,0x00,0xf8]
-; GFX11-NEXT: v_mov_b32_e32 v0, 0xffff8000 ; encoding: [0xff,0x02,0x00,0x7e,0x00,0x80,0xff,0xff]
-; GFX11-NEXT: s_mov_b32 s3, 0x31016000 ; encoding: [0xff,0x00,0x83,0xbe,0x00,0x60,0x01,0x31]
-; GFX11-NEXT: s_mov_b32 s2, -1 ; encoding: [0xc1,0x00,0x82,0xbe]
-; GFX11-NEXT: s_waitcnt lgkmcnt(0) ; encoding: [0x07,0xfc,0x89,0xbf]
-; GFX11-NEXT: buffer_store_b16 v0, off, s[0:3], 0 ; encoding: [0x00,0x00,0x64,0xe0,0x00,0x00,0x00,0x80]
-; GFX11-NEXT: s_endpgm ; encoding: [0x00,0x00,0xb0,0xbf]
+; GFX11-TRUE16-LABEL: store_imm_neg_0.0_f16:
+; GFX11-TRUE16: ; %bb.0:
+; GFX11-TRUE16-NEXT: s_load_b64 s[0:1], s[4:5], 0x0 ; encoding: [0x02,0x00,0x04,0xf4,0x00,0x00,0x00,0xf8]
+; GFX11-TRUE16-NEXT: v_mov_b16_e32 v0.l, 0x8000 ; encoding: [0xff,0x38,0x00,0x7e,0x00,0x80,0xff,0xff]
+; GFX11-TRUE16-NEXT: s_mov_b32 s3, 0x31016000 ; encoding: [0xff,0x00,0x83,0xbe,0x00,0x60,0x01,0x31]
+; GFX11-TRUE16-NEXT: s_mov_b32 s2, -1 ; encoding: [0xc1,0x00,0x82,0xbe]
+; GFX11-TRUE16-NEXT: s_waitcnt lgkmcnt(0) ; encoding: [0x07,0xfc,0x89,0xbf]
+; GFX11-TRUE16-NEXT: buffer_store_b16 v0, off, s[0:3], 0 ; encoding: [0x00,0x00,0x64,0xe0,0x00,0x00,0x00,0x80]
+; GFX11-TRUE16-NEXT: s_endpgm ; encoding: [0x00,0x00,0xb0,0xbf]
+;
+; GFX11-FAKE16-LABEL: store_imm_neg_0.0_f16:
+; GFX11-FAKE16: ; %bb.0:
+; GFX11-FAKE16-NEXT: s_load_b64 s[0:1], s[4:5], 0x0 ; encoding: [0x02,0x00,0x04,0xf4,0x00,0x00,0x00,0xf8]
+; GFX11-FAKE16-NEXT: v_mov_b32_e32 v0, 0xffff8000 ; encoding: [0xff,0x02,0x00,0x7e,0x00,0x80,0xff,0xff]
+; GFX11-FAKE16-NEXT: s_mov_b32 s3, 0x31016000 ; encoding: [0xff,0x00,0x83,0xbe,0x00,0x60,0x01,0x31]
+; GFX11-FAKE16-NEXT: s_mov_b32 s2, -1 ; encoding: [0xc1,0x00,0x82,0xbe]
+; GFX11-FAKE16-NEXT: s_waitcnt lgkmcnt(0) ; encoding: [0x07,0xfc,0x89,0xbf]
+; GFX11-FAKE16-NEXT: buffer_store_b16 v0, off, s[0:3], 0 ; encoding: [0x00,0x00,0x64,0xe0,0x00,0x00,0x00,0x80]
+; GFX11-FAKE16-NEXT: s_endpgm ; encoding: [0x00,0x00,0xb0,0xbf]
;
; VI-LABEL: store_imm_neg_0.0_f16:
; VI: ; %bb.0:
@@ -154,15 +185,25 @@ define amdgpu_kernel void @store_inline_imm_0.5_f16(ptr addrspace(1) %out) {
; GFX10-NEXT: buffer_store_short v0, off, s[0:3], 0 ; encoding: [0x00,0x00,0x68,0xe0,0x00,0x00,0x00,0x80]
; GFX10-NEXT: s_endpgm ; encoding: [0x00,0x00,0x81,0xbf]
;
-; GFX11-LABEL: store_inline_imm_0.5_f16:
-; GFX11: ; %bb.0:
-; GFX11-NEXT: s_load_b64 s[0:1], s[4:5], 0x0 ; encoding: [0x02,0x00,0x04,0xf4,0x00,0x00,0x00,0xf8]
-; GFX11-NEXT: v_mov_b32_e32 v0, 0x3800 ; encoding: [0xff,0x02,0x00,0x7e,0x00,0x38,0x00,0x00]
-; GFX11-NEXT: s_mov_b32 s3, 0x31016000 ; encoding: [0xff,0x00,0x83,0xbe,0x00,0x60,0x01,0x31]
-; GFX11-NEXT: s_mov_b32 s2, -1 ; encoding: [0xc1,0x00,0x82,0xbe]
-; GFX11-NEXT: s_waitcnt lgkmcnt(0) ; encoding: [0x07,0xfc,0x89,0xbf]
-; GFX11-NEXT: buffer_store_b16 v0, off, s[0:3], 0 ; encoding: [0x00,0x00,0x64,0xe0,0x00,0x00,0x00,0x80]
-; GFX11-NEXT: s_endpgm ; encoding: [0x00,0x00,0xb0,0xbf]
+; GFX11-TRUE16-LABEL: store_inline_imm_0.5_f16:
+; GFX11-TRUE16: ; %bb.0:
+; GFX11-TRUE16-NEXT: s_load_b64 s[0:1], s[4:5], 0x0 ; encoding: [0x02,0x00,0x04,0xf4,0x00,0x00,0x00,0xf8]
+; GFX11-TRUE16-NEXT: v_mov_b16_e32 v0.l, 0x3800 ; encoding: [0xff,0x38,0x00,0x7e,0x00,0x38,0x00,0x00]
+; GFX11-TRUE16-NEXT: s_mov_b32 s3, 0x31016000 ; encoding: [0xff,0x00,0x83,0xbe,0x00,0x60,0x01,0x31]
+; GFX11-TRUE16-NEXT: s_mov_b32 s2, -1 ; encoding: [0xc1,0x00,0x82,0xbe]
+; GFX11-TRUE16-NEXT: s_waitcnt lgkmcnt(0) ; encoding: [0x07,0xfc,0x89,0xbf]
+; GFX11-TRUE16-NEXT: buffer_store_b16 v0, off, s[0:3], 0 ; encoding: [0x00,0x00,0x64,0xe0,0x00,0x00,0x00,0x80]
+; GFX11-TRUE16-NEXT: s_endpgm ; encoding: [0x00,0x00,0xb0,0xbf]
+;
+; GFX11-FAKE16-LABEL: store_inline_imm_0.5_f16:
+; GFX11-FAKE16: ; %bb.0:
+; GFX11-FAKE16-NEXT: s_load_b64 s[0:1], s[4:5], 0x0 ; encoding: [0x02,0x00,0x04,0xf4,0x00,0x00,0x00,0xf8]
+; GFX11-FAKE16-NEXT: v_mov_b32_e32 v0, 0x3800 ; encoding: [0xff,0x02,0x00,0x7e,0x00,0x38,0x00,0x00]
+; GFX11-FAKE16-NEXT: s_mov_b32 s3, 0x31016000 ; encoding: [0xff,0x00,0x83,0xbe,0x00,0x60,0x01,0x31]
+; GFX11-FAKE16-NEXT: s_mov_b32 s2, -1 ; encoding: [0xc1,0x00,0x82,0xbe]
+; GFX11-FAKE16-NEXT: s_waitcnt lgkmcnt(0) ; encoding: [0x07,0xfc,0x89,0xbf]
+; GFX11-FAKE16-NEXT: buffer_store_b16 v0, off, s[0:3], 0 ; encoding: [0x00,0x00,0x64,0xe0,0x00,0x00,0x00,0x80]
+; GFX11-FAKE16-NEXT: s_endpgm ; encoding: [0x00,0x00,0xb0,0xbf]
;
; VI-LABEL: store_inline_imm_0.5_f16:
; VI: ; %bb.0:
@@ -198,15 +239,25 @@ define amdgpu_kernel void @store_inline_imm_m_0.5_f16(ptr addrspace(1) %out) {
; GFX10-NEXT: buffer_store_short v0, off, s[0:3], 0 ; encoding: [0x00,0x00,0x68,0xe0,0x00,0x00,0x00,0x80]
; GFX10-NEXT: s_endpgm ; encoding: [0x00,0x00,0x81,0xbf]
;
-; GFX11-LABEL: store_inline_imm_m_0.5_f16:
-; GFX11: ; %bb.0:
-; GFX11-NEXT: s_load_b64 s[0:1], s[4:5], 0x0 ; encoding: [0x02,0x00,0x04,0xf4,0x00,0x00,0x00,0xf8]
-; GFX11-NEXT: v_mov_b32_e32 v0, 0xffffb800 ; encoding: [0xff,0x02,0x00,0x7e,0x00,0xb8,0xff,0xff]
-; GFX11-NEXT: s_mov_b32 s3, 0x31016000 ; encoding: [0xff,0x00,0x83,0xbe,0x00,0x60,0x01,0x31]
-; GFX11-NEXT: s_mov_b32 s2, -1 ; encoding: [0xc1,0x00,0x82,0xbe]
-; GFX11-NEXT: s_waitcnt lgkmcnt(0) ; encoding: [0x07,0xfc,0x89,0xbf]
-; GFX11-NEXT: buffer_store_b16 v0, off, s[0:3], 0 ; encoding: [0x00,0x00,0x64,0xe0,0x00,0x00,0x00,0x80]
-; GFX11-NEXT: s_endpgm ; encoding: [0x00,0x00,0xb0,0xbf]
+; GFX11-TRUE16-LABEL: store_inline_imm_m_0.5_f16:
+; GFX11-TRUE16: ; %bb.0:
+; GFX11-TRUE16-NEXT: s_load_b64 s[0:1], s[4:5], 0x0 ; encoding: [0x02,0x00,0x04,0xf4,0x00,0x00,0x00,0xf8]
+; GFX11-TRUE16-NEXT: v_mov_b16_e32 v0.l, 0xb800 ; encoding: [0xff,0x38,0x00,0x7e,0x00,0xb8,0xff,0xff]
+; GFX11-TRUE16-NEXT: s_mov_b32 s3, 0x31016000 ; encoding: [0xff,0x00,0x83,0xbe,0x00,0x60,0x01,0x31]
+; GFX11-TRUE16-NEXT: s_mov_b32 s2, -1 ; encoding: [0xc1,0x00,0x82,0xbe]
+; GFX11-TRUE16-NEXT: s_waitcnt lgkmcnt(0) ; encoding: [0x07,0xfc,0x89,0xbf]
+; GFX11-TRUE16-NEXT: buffer_store_b16 v0, off, s[0:3], 0 ; encoding: [0x00,0x00,0x64,0xe0,0x00,0x00,0x00,0x80]
+; GFX11-TRUE16-NEXT: s_endpgm ; encoding: [0x00,0x00,0xb0,0xbf]
+;
+; GFX11-FAKE16-LABEL: store_inline_imm_m_0.5_f16:
+; GFX11-FAKE16: ; %bb.0:
+; GFX11-FAKE16-NEXT: s_load_b64 s[0:1], s[4:5], 0x0 ; encoding: [0x02,0x00,0x04,0xf4,0x00,0x00,0x00,0xf8]
+; GFX11-FAKE16-NEXT: v_mov_b32_e32 v0, 0xffffb800 ; encoding: [0xff,0x02,0x00,0x7e,0x00,0xb8,0xff,0xff]
+; GFX11-FAKE16-NEXT: s_mov_b32 s3, 0x31016000 ; encoding: [0xff,0x00,0x83,0xbe,0x00,0x60,0x01,0x31]
+; GFX11-FAKE16-NEXT: s_mov_b32 s2, -1 ; encoding: [0xc1,0x00,0x82,0xbe]
+; GFX11-FAKE16-NEXT: s_waitcnt lgkmcnt(0) ; encoding: [0x07,0xfc,0x89,0xbf]
+; GFX11-FAKE16-NEXT: buffer_store_b16 v0, off, s[0:3], 0 ; encoding: [0x00,0x00,0x64,0xe0,0x00,0x00,0x00,0x80]
+; GFX11-FAKE16-NEXT: s_endpgm ; encoding: [0x00,0x00,0xb0,0xbf]
;
; VI-LABEL: store_inline_imm_m_0.5_f16:
; VI: ; %bb.0:
@@ -242,15 +293,25 @@ define amdgpu_kernel void @store_inline_imm_1.0_f16(ptr addrspace(1) %out) {
; GFX10-NEXT: buffer_store_short v0, off, s[0:3], 0 ; encoding: [0x00,0x00,0x68,0xe0,0x00,0x00,0x00,0x80]
; GFX10-NEXT: s_endpgm ; encoding: [0x00,0x00,0x81,0xbf]
;
-; GFX11-LABEL: store_inline_imm_1.0_f16:
-; GFX11: ; %bb.0:
-; GFX11-NEXT: s_load_b64 s[0:1], s[4:5], 0x0 ; encoding: [0x02,0x00,0x04,0xf4,0x00,0x00,0x00,0xf8]
-; GFX11-NEXT: v_mov_b32_e32 v0, 0x3c00 ; encoding: [0xff,0x02,0x00,0x7e,0x00,0x3c,0x00,0x00]
-; GFX11-NEXT: s_mov_b32 s3, 0x31016000 ; encoding: [0xff,0x00,0x83,0xbe,0x00,0x60,0x01,0x31]
-; GFX11-NEXT: s_mov_b32 s2, -1 ; encoding: [0xc1,0x00,0x82,0xbe]
-; GFX11-NEXT: s_waitcnt lgkmcnt(0) ; encoding: [0x07,0xfc,0x89,0xbf]
-; GFX11-NEXT: buffer_store_b16 v0, off, s[0:3], 0 ; encoding: [0x00,0x00,0x64,0xe0,0x00,0x00,0x00,0x80]
-; GFX11-NEXT: s_endpgm ; encoding: [0x00,0x00,0xb0,0xbf]
+; GFX11-TRUE16-LABEL: store_inline_imm_1.0_f16:
+; GFX11-TRUE16: ; %bb.0:
+; GFX11-TRUE16-NEXT: s_load_b64 s[0:1], s[4:5], 0x0 ; encoding: [0x02,0x00,0x04,0xf4,0x00,0x00,0x00,0xf8]
+; GFX11-TRUE16-NEXT: v_mov_b16_e32 v0.l, 0x3c00 ; encoding: [0xff,0x38,0x00,0x7e,0x00,0x3c,0x00,0x00]
+; GFX11-TRUE16-NEXT: s_mov_b32 s3, 0x31016000 ; encoding: [0xff,0x00,0x83,0xbe,0x00,0x60,0x01,0x31]
+; GFX11-TRUE16-NEXT: s_mov_b32 s2, -1 ; encoding: [0xc1,0x00,0x82,0xbe]
+; GFX11-TRUE16-NEXT: s_waitcnt lgkmcnt(0) ; encoding: [0x07,0xfc,0x89,0xbf]
+; GFX11-TRUE16-NEXT: buffer_store_b16 v0, off, s[0:3], 0 ; encoding: [0x00,0x00,0x64,0xe0,0x00,0x00,0x00,0x80]
+; GFX11-TRUE16-NEXT: s_endpgm ; encoding: [0x00,0x00,0xb0,0xbf]
+;
+; GFX11-FAKE16-LABEL: store_inline_imm_1.0_f16:
+; GFX11-FAKE16: ; %bb.0:
+; GFX11-FAKE16-NEXT: s_load_b64 s[0:1], s[4:5], 0x0 ; encoding: [0x02,0x00,0x04,0xf4,0x00,0x00,0x00,0xf8]
+; GFX11-FAKE16-NEXT: v_mov_b32_e32 v0, 0x3c00 ; encoding: [0xff,0x02,0x00,0x7e,0x00,0x3c,0x00,0x00]
+; GFX11-FAKE16-NEXT: s_mov_b32 s3, 0x31016000 ; encoding: [0xff,0x00,0x83,0xbe,0x00,0x60,0x01,0x31]
+; GFX11-FAKE16-NEXT: s_mov_b32 s2, -1 ; encoding: [0xc1,0x00,0x82,0xbe]
+; GFX11-FAKE16-NEXT: s_waitcnt lgkmcnt(0) ; encoding: [0x07,0xfc,0x89,0xbf]
+; GFX11-FAKE16-NEXT: buffer_store_b16 v0, off, s[0:3], 0 ; encoding: [0x00,0x00,0x64,0xe0,0x00,0x00,0x00,0x80]
+; GFX11-FAKE16-NEXT: s_endpgm ; encoding: [0x00,0x00,0xb0,0xbf]
;
; VI-LABEL: store_inline_imm_1.0_f16:
; VI: ; %bb.0:
@@ -286,15 +347,25 @@ define amdgpu_kernel void @store_inline_imm_m_1.0_f16(ptr addrspace(1) %out) {
; GFX10-NEXT: buffer_store_short v0, off, s[0:3], 0 ; encoding: [0x00,0x00,0x68,0xe0,0x00,0x00,0x00,0x80]
; GFX10-NEXT: s_endpgm ; encoding: [0x00,0x00,0x81,0xbf]
;
-; GFX11-LABEL: store_inline_imm_m_1.0_f16:
-; GFX11: ; %bb.0:
-; GFX11-NEXT: s_load_b64 s[0:1], s[4:5], 0x0 ; encoding: [0x02,0x00,0x04,0xf4,0x00,0x00,0x00,0xf8]
-; GFX11-NEXT: v_mov_b32_e32 v0, 0xffffbc00 ; encoding: [0xff,0x02,0x00,0x7e,0x00,0xbc,0xff,0xff]
-; GFX11-NEXT: s_mov_b32 s3, 0x31016000 ; encoding: [0xff,0x00,0x83,0xbe,0x00,0x60,0x01,0x31]
-; GFX11-NEXT: s_mov_b32 s2, -1 ; encoding: [0xc1,0x00,0x82,0xbe]
-; GFX11-NEXT: s_waitcnt lgkmcnt(0) ; encoding: [0x07,0xfc,0x89,0xbf]
-; GFX11-NEXT: buffer_store_b16 v0, off, s[0:3], 0 ; encoding: [0x00,0x00,0x64,0xe0,0x00,0x00,0x00,0x80]
-; GFX11-NEXT: s_endpgm ; encoding: [0x00,0x00,0xb0,0xbf]
+; GFX11-TRUE16-LABEL: store_inline_imm_m_1.0_f16:
+; GFX11-TRUE16: ; %bb.0:
+; GFX11-TRUE16-NEXT: s_load_b64 s[0:1], s[4:5], 0x0 ; encoding: [0x02,0x00,0x04,0xf4,0x00,0x00,0x00,0xf8]
+; GFX11-TRUE16-NEXT: v_mov_b16_e32 v0.l, 0xbc00 ; encoding: [0xff,0x38,0x00,0x7e,0x00,0xbc,0xff,0xff]
+; GFX11-TRUE16-NEXT: s_mov_b32 s3, 0x31016000 ; encoding: [0xff,0x00,0x83,0xbe,0x00,0x60,0x01,0x31]
+; GFX11-TRUE16-NEXT: s_mov_b32 s2, -1 ; encoding: [0xc1,0x00,0x82,0xbe]
+; GFX11-TRUE16-NEXT: s_waitcnt lgkmcnt(0) ; encoding: [0x07,0xfc,0x89,0xbf]
+; GFX11-TRUE16-NEXT: buffer_store_b16 v0, off, s[0:3], 0 ; encoding: [0x00,0x00,0x64,0xe0,0x00,0x00,0x00,0x80]
+; GFX11-TRUE16-NEXT: s_endpgm ; encoding: [0x00,0x00,0xb0,0xbf]
+;
+; GFX11-FAKE16-LABEL: store_inline_imm_m_1.0_f16:
+; GFX11-FAKE16: ; %bb.0:
+; GFX11-FAKE16-NEXT: s_load_b64 s[0:1], s[4:5], 0x0 ; encoding: [0x02,0x00,0x04,0xf4,0x00,0x00,0x00,0xf8]
+; GFX11-FAKE16-NEXT: v_mov_b32_e32 v0, 0xffffbc00 ; encoding: [0xff,0x02,0x00,0x7e,0x00,0xbc,0xff,0xff]
+; GFX11-FAKE16-NEXT: s_mov_b32 s3, 0x31016000 ; encoding: [0xff,0x00,0x83,0xbe,0x00,0x60,0x01,0x31]
+; GFX11-FAKE16-NEXT: s_mov_b32 s2, -1 ; encoding: [0xc1,0x00,0x82,0xbe]
+; GFX11-FAKE16-NEXT: s_waitcnt lgkmcnt(0) ; encoding: [0x07,0xfc,0x89,0xbf]
+; GFX11-FAKE16-NEXT: buffer_store_b16 v0, off, s[0:3], 0 ; encoding: [0x00,0x00,0x64,0xe0,0x00,0x00,0x00,0x80]
+; GFX11-FAKE16-NEXT: s_endpgm ; encoding: [0x00,0x00,0xb0,0xbf]
;
; VI-LABEL: store_inline_imm_m_1.0_f16:
; VI: ; %bb.0:
@@ -330,15 +401,25 @@ define amdgpu_kernel void @store_inline_imm_2.0_f16(ptr addrspace(1) %out) {
; GFX10-NEXT: buffer_store_short v0, off, s[0:3], 0 ; encoding: [0x00,0x00,0x68,0xe0,0x00,0x00,0x00,0x80]
; GFX10-NEXT: s_endpgm ; encoding: [0x00,0x00,0x81,0xbf]
;
-; GFX11-LABEL: store_inline_imm_2.0_f16:
-; GFX11: ; %bb.0:
-; GFX11-NEXT: s_load_b64 s[0:1], s[4:5], 0x0 ; encoding: [0x02,0x00,0x04,0xf4,0x00,0x00,0x00,0xf8]
-; GFX11-NEXT: v_mov_b32_e32 v0, 0x4000 ; encoding: [0xff,0x02,0x00,0x7e,0x00,0x40,0x00,0x00]
-; GFX11-NEXT: s_mov_b32 s3, 0x31016000 ; encoding: [0xff,0x00,0x83,0xbe,0x00,0x60,0x01,0x31]
-; GFX11-NEXT: s_mov_b32 s2, -1 ; encoding: [0xc1,0x00,0x82,0xbe]
-; GFX11-NEXT: s_waitcnt lgkmcnt(0) ; encoding: [0x07,0xfc,0x89,0xbf]
-; GFX11-NEXT: buffer_store_b16 v0, off, s[0:3], 0 ; encoding: [0x00,0x00,0x64,0xe0,0x00,0x00,0x00,0x80]
-; GFX11-NEXT: s_endpgm ; encoding: [0x00,0x00,0xb0,0xbf]
+; GFX11-TRUE16-LABEL: store_inline_imm_2.0_f16:
+; GFX11-TRUE16: ; %bb.0:
+; GFX11-TRUE16-NEXT: s_load_b64 s[0:1], s[4:5], 0x0 ; encoding: [0x02,0x00,0x04,0xf4,0x00,0x00,0x00,0xf8]
+; GFX11-TRUE16-NEXT: v_mov_b16_e32 v0.l, 0x4000 ; encoding: [0xff,0x38,0x00,0x7e,0x00,0x40,0x00,0x00]
+; GFX11-TRUE16-NEXT: s_mov_b32 s3, 0x31016000 ; encoding: [0xff,0x00,0x83,0xbe,0x00,0x60,0x01,0x31]
+; GFX11-TRUE16-NEXT: s_mov_b32 s2, -1 ; encoding: [0xc1,0x00,0x82,0xbe]
+; GFX11-TRUE16-NEXT: s_waitcnt lgkmcnt(0) ; encoding: [0x07,0xfc,0x89,0xbf]
+; GFX11-TRUE16-NEXT: buffer_store_b16 v0, off, s[0:3], 0 ; encoding: [0x00,0x00,0x64,0xe0,0x00,0x00,0x00,0x80]
+; GFX11-TRUE16-NEXT: s_endpgm ; encoding: [0x00,0x00,0xb0,0xbf]
+;
+; GFX11-FAKE16-LABEL: store_inline_imm_2.0_f16:
+; GFX11-FAKE16: ; %bb.0:
+; GFX11-FAKE16-NEXT: s_load_b64 s[0:1], s[4:5], 0x0 ; encoding: [0x02,0x00,0x04,0xf4,0x00,0x00,0x00,0xf8]
+; GFX11-FAKE16-NEXT: v_mov_b32_e32 v0, 0x4000 ; encoding: [0xff,0x02,0x00,0x7e,0x00,0x40,0x00,0x00]
+; GFX11-FAKE16-NEXT: s_mov_b32 s3, 0x31016000 ; encoding: [0xff,0x00,0x83,0xbe,0x00,0x60,0x01,0x31]
+; GFX11-FAKE16-NEXT: s_mov_b32 s2, -1 ; encoding: [0xc1,0x00,0x82,0xbe]
+; GFX11-FAKE16-NEXT: s_waitcnt lgkmcnt(0) ; encoding: [0x07,0xfc,0x89,0xbf]
+; GFX11-FAKE16-NEXT: buffer_store_b16 v0, off, s[0:3], 0 ; encoding: [0x00,0x00,0x64,0xe0,0x00,0x00,0x00,0x80]
+; GFX11-FAKE16-NEXT: s_endpgm ; encoding: [0x00,0x00,0xb0,0xbf]
;
; VI-LABEL: store_inline_imm_2.0_f16:
; VI: ; %bb.0:
@@ -374,15 +455,25 @@ define amdgpu_kernel void @store_inline_imm_m_2.0_f16(ptr addrspace(1) %out) {
; GFX10-NEXT: buffer_store_short v0, off, s[0:3], 0 ; encoding: [0x00,0x00,0x68,0xe0,0x00,0x00,0x00,0x80]
; GFX10-NEXT: s_endpgm ; encoding: [0x00,0x00,0x81,0xbf]
;
-; GFX11-LABEL: store_inline_imm_m_2.0_f16:
-; GFX11: ; %bb.0:
-; GFX11-NEXT: s_load_b64 s[0:1], s[4:5], 0x0 ; encoding: [0x02,0x00,0x04,0xf4,0x00,0x00,0x00,0xf8]
-; GFX11-NEXT: v_mov_b32_e32 v0, 0xffffc000 ; encoding: [0xff,0x02,0x00,0x7e,0x00,0xc0,0xff,0xff]
-; GFX11-NEXT: s_mov_b32 s3, 0x31016000 ; encoding: [0xff,0x00,0x83,0xbe,0x00,0x60,0x01,0x31]
-; GFX11-NEXT: s_mov_b32 s2, -1 ; encoding: [0xc1,0x00,0x82,0xbe]
-; GFX11-NEXT: s_waitcnt lgkmcnt(0) ; encoding: [0x07,0xfc,0x89,0xbf]
-; GFX11-NEXT: buffer_store_b16 v0, off, s[0:3], 0 ; encoding: [0x00,0x00,0x64,0xe0,0x00,0x00,0x00,0x80]
-; GFX11-NEXT: s_endpgm ; encoding: [0x00,0x00,0xb0,0xbf]
+; GFX11-TRUE16-LABEL: store_inline_imm_m_2.0_f16:
+; GFX11-TRUE16: ; %bb.0:
+; GFX11-TRUE16-NEXT: s_load_b64 s[0:1], s[4:5], 0x0 ; encoding: [0x02,0x00,0x04,0xf4,0x00,0x00,0x00,0xf8]
+; GFX11-TRUE16-NEXT: v_mov_b16_e32 v0.l, 0xc000 ; encoding: [0xff,0x38,0x00,0x7e,0x00,0xc0,0xff,0xff]
+; GFX11-TRUE16-NEXT: s_mov_b32 s3, 0x31016000 ; encoding: [0xff,0x00,0x83,0xbe,0x00,0x60,0x01,0x31]
+; GFX11-TRUE16-NEXT: s_mov_b32 s2, -1 ; encoding: [0xc1,0x00,0x82,0xbe]
+; GFX11-TRUE16-NEXT: s_waitcnt lgkmcnt(0) ; encoding: [0x07,0xfc,0x89,0xbf]
+; GFX11-TRUE16-NEXT: buffer_store_b16 v0, off, s[0:3], 0 ; encoding: [0x00,0x00,0x64,0xe0,0x00,0x00,0x00,0x80]
+; GFX11-TRUE16-NEXT: s_endpgm ; encoding: [0x00,0x00,0xb0,0xbf]
+;
+; GFX11-FAKE16-LABEL: store_inline_imm_m_2.0_f16:
+; GFX11-FAKE16: ; %bb.0:
+; GFX11-FAKE16-NEXT: s_load_b64 s[0:1], s[4:5], 0x0 ; encoding: [0x02,0x00,0x04,0xf4,0x00,0x00,0x00,0xf8]
+; GFX11-FAKE16-NEXT: v_mov_b32_e32 v0, 0xffffc000 ; encoding: [0xff,0x02,0x00,0x7e,0x00,0xc0,0xff,0xff]
+; GFX11-FAKE16-NEXT: s_mov_b32 s3, 0x31016000 ; encoding: [0xff,0x00,0x83,0xbe,0x00,0x60,0x01,0x31]
+; GFX11-FAKE16-NEXT: s_mov_b32 s2, -1 ; encoding: [0xc1,0x00,0x82,0xbe]
+; GFX11-FAKE16-NEXT: s_waitcnt lgkmcnt(0) ; encoding: [0x07,0xfc,0x89,0xbf]
+; GFX11-FAKE16-NEXT: buffer_store_b16 v0, off, s[0:3], 0 ; encoding: [0x00,0x00,0x64,0xe0,0x00,0x00,0x00,0x80]
+; GFX11-FAKE16-NEXT: s_endpgm ; encoding: [0x00,0x00,0xb0,0xbf]
;
; VI-LABEL: store_inline_imm_m_2.0_f16:
; VI: ; %bb.0:
@@ -418,15 +509,25 @@ define amdgpu_kernel void @store_inline_imm_4.0_f16(ptr addrspace(1) %out) {
; GFX10-NEXT: buffer_store_short v0, off, s[0:3], 0 ; encoding: [0x00,0x00,0x68,0xe0,0x00,0x00,0x00,0x80]
; GFX10-NEXT: s_endpgm ; encoding: [0x00,0x00,0x81,0xbf]
;
-; GFX11-LABEL: store_inline_imm_4.0_f16:
-; GFX11: ; %bb.0:
-; GFX11-NEXT: s_load_b64 s[0:1], s[4:5], 0x0 ; encoding: [0x02,0x00,0x04,0xf4,0x00,0x00,0x00,0xf8]
-; GFX11-NEXT: v_mov_b32_e32 v0, 0x4400 ; encoding: [0xff,0x02,0x00,0x7e,0x00,0x44,0x00,0x00]
-; GFX11-NEXT: s_mov_b32 s3, 0x31016000 ; encoding: [0xff,0x00,0x83,0xbe,0x00,0x60,0x01,0x31]
-; GFX11-NEXT: s_mov_b32 s2, -1 ; encoding: [0xc1,0x00,0x82,0xbe]
-; GFX11-NEXT: s_waitcnt lgkmcnt(0) ; encoding: [0x07,0xfc,0x89,0xbf]
-; GFX11-NEXT: buffer_store_b16 v0, off, s[0:3], 0 ; encoding: [0x00,0x00,0x64,0xe0,0x00,0x00,0x00,0x80]
-; GFX11-NEXT: s_endpgm ; encoding: [0x00,0x00,0xb0,0xbf]
+; GFX11-TRUE16-LABEL: store_inline_imm_4.0_f16:
+; GFX11-TRUE16: ; %bb.0:
+; GFX11-TRUE16-NEXT: s_load_b64 s[0:1], s[4:5], 0x0 ; encoding: [0x02,0x00,0x04,0xf4,0x00,0x00,0x00,0xf8]
+; GFX11-TRUE16-NEXT: v_mov_b16_e32 v0.l, 0x4400 ; encoding: [0xff,0x38,0x00,0x7e,0x00,0x44,0x00,0x00]
+; GFX11-TRUE16-NEXT: s_mov_b32 s3, 0x31016000 ; encoding: [0xff,0x00,0x83,0xbe,0x00,0x60,0x01,0x31]
+; GFX11-TRUE16-NEXT: s_mov_b32 s2, -1 ; encoding: [0xc1,0x00,0x82,0xbe]
+; GFX11-TRUE16-NEXT: s_waitcnt lgkmcnt(0) ; encoding: [0x07,0xfc,0x89,0xbf]
+; GFX11-TRUE16-NEXT: buffer_store_b16 v0, off, s[0:3], 0 ; encoding: [0x00,0x00,0x64,0xe0,0x00,0x00,0x00,0x80]
+; GFX11-TRUE16-NEXT: s_endpgm ; encoding: [0x00,0x00,0xb0,0xbf]
+;
+; GFX11-FAKE16-LABEL: store_inline_imm_4.0_f16:
+; GFX11-FAKE16: ; %bb.0:
+; GFX11-FAKE16-NEXT: s_load_b64 s[0:1], s[4:5], 0x0 ; encoding: [0x02,0x00,0x04,0xf4,0x00,0x00,0x00,0xf8]
+; GFX11-FAKE16-NEXT: v_mov_b32_e32 v0, 0x4400 ; encoding: [0xff,0x02,0x00,0x7e,0x00,0x44,0x00,0x00]
+; GFX11-FAKE16-NEXT: s_mov_b32 s3, 0x31016000 ; encoding: [0xff,0x00,0x83,0xbe,0x00,0x60,0x01,0x31]
+; GFX11-FAKE16-NEXT: s_mov_b32 s2, -1 ; encoding: [0xc1,0x00,0x82,0xbe]
+; GFX11-FAKE16-NEXT: s_waitcnt lgkmcnt(0) ; encoding: [0x07,0xfc,0x89,0xbf]
+; GFX11-FAKE16-NEXT: buffer_store_b16 v0, off, s[0:3], 0 ; encoding: [0x00,0x00,0x64,0xe0,0x00,0x00,0x00,0x80]
+; GFX11-FAKE16-NEXT: s_endpgm ; encoding: [0x00,0x00,0xb0,0xbf]
;
; VI-LABEL: store_inline_imm_4.0_f16:
; VI: ; %bb.0:
@@ -462,15 +563,25 @@ define amdgpu_kernel void @store_inline_imm_m_4.0_f16(ptr addrspace(1) %out) {
; GFX10-NEXT: buffer_store_short v0, off, s[0:3], 0 ; encoding: [0x00,0x00,0x68,0xe0,0x00,0x00,0x00,0x80]
; GFX10-NEXT: s_endpgm ; encoding: [0x00,0x00,0x81,0xbf]
;
-; GFX11-LABEL: store_inline_imm_m_4.0_f16:
-; GFX11: ; %bb.0:
-; GFX11-NEXT: s_load_b64 s[0:1], s[4:5], 0x0 ; encoding: [0x02,0x00,0x04,0xf4,0x00,0x00,0x00,0xf8]
-; GFX11-NEXT: v_mov_b32_e32 v0, 0xffffc400 ; encoding: [0xff,0x02,0x00,0x7e,0x00,0xc4,0xff,0xff]
-; GFX11-NEXT: s_mov_b32 s3, 0x31016000 ; encoding: [0xff,0x00,0x83,0xbe,0x00,0x60,0x01,0x31]
-; GFX11-NEXT: s_mov_b32 s2, -1 ; encoding: [0xc1,0x00,0x82,0xbe]
-; GFX11-NEXT: s_waitcnt lgkmcnt(0) ; encoding: [0x07,0xfc,0x89,0xbf]
-; GFX11-NEXT: buffer_store_b16 v0, off, s[0:3], 0 ; encoding: [0x00,0x00,0x64,0xe0,0x00,0x00,0x00,0x80]
-; GFX11-NEXT: s_endpgm ; encoding: [0x00,0x00,0xb0,0xbf]
+; GFX11-TRUE16-LABEL: store_inline_imm_m_4.0_f16:
+; GFX11-TRUE16: ; %bb.0:
+; GFX11-TRUE16-NEXT: s_load_b64 s[0:1], s[4:5], 0x0 ; encoding: [0x02,0x00,0x04,0xf4,0x00,0x00,0x00,0xf8]
+; GFX11-TRUE16-NEXT: v_mov_b16_e32 v0.l, 0xc400 ; encoding: [0xff,0x38,0x00,0x7e,0x00,0xc4,0xff,0xff]
+; GFX11-TRUE16-NEXT: s_mov_b32 s3, 0x31016000 ; encoding: [0xff,0x00,0x83,0xbe,0x00,0x60,0x01,0x31]
+; GFX11-TRUE16-NEXT: s_mov_b32 s2, -1 ; encoding: [0xc1,0x00,0x82,0xbe]
+; GFX11-TRUE16-NEXT: s_waitcnt lgkmcnt(0) ; encoding: [0x07,0xfc,0x89,0xbf]
+; GFX11-TRUE16-NEXT: buffer_store_b16 v0, off, s[0:3], 0 ; encoding: [0x00,0x00,0x64,0xe0,0x00,0x00,0x00,0x80]
+; GFX11-TRUE16-NEXT: s_endpgm ; encoding: [0x00,0x00,0xb0,0xbf]
+;
+; GFX11-FAKE16-LABEL: store_inline_imm_m_4.0_f16:
+; GFX11-FAKE16: ; %bb.0:
+; GFX11-FAKE16-NEXT: s_load_b64 s[0:1], s[4:5], 0x0 ; encoding: [0x02,0x00,0x04,0xf4,0x00,0x00,0x00,0xf8]
+; GFX11-FAKE16-NEXT: v_mov_b32_e32 v0, 0xffffc400 ; encoding: [0xff,0x02,0x00,0x7e,0x00,0xc4,0xff,0xff]
+; GFX11-FAKE16-NEXT: s_mov_b32 s3, 0x31016000 ; encoding: [0xff,0x00,0x83,0xbe,0x00,0x60,0x01,0x31]
+; GFX11-FAKE16-NEXT: s_mov_b32 s2, -1 ; encoding: [0xc1,0x00,0x82,0xbe]
+; GFX11-FAKE16-NEXT: s_waitcnt lgkmcnt(0) ; encoding: [0x07,0xfc,0x89,0xbf]
+; GFX11-FAKE16-NEXT: buffer_store_b16 v0, off, s[0:3], 0 ; encoding: [0x00,0x00,0x64,0xe0,0x00,0x00,0x00,0x80]
+; GFX11-FAKE16-NEXT: s_endpgm ; encoding: [0x00,0x00,0xb0,0xbf]
;
; VI-LABEL: store_inline_imm_m_4.0_f16:
; VI: ; %bb.0:
@@ -506,15 +617,25 @@ define amdgpu_kernel void @store_inline_imm_inv_2pi_f16(ptr addrspace(1) %out) {
; GFX10-NEXT: buffer_store_short v0, off, s[0:3], 0 ; encoding: [0x00,0x00,0x68,0xe0,0x00,0x00,0x00,0x80]
; GFX10-NEXT: s_endpgm ; encoding: [0x00,0x00,0x81,0xbf]
;
-; GFX11-LABEL: store_inline_imm_inv_2pi_f16:
-; GFX11: ; %bb.0:
-; GFX11-NEXT: s_load_b64 s[0:1], s[4:5], 0x0 ; encoding: [0x02,0x00,0x04,0xf4,0x00,0x00,0x00,0xf8]
-; GFX11-NEXT: v_mov_b32_e32 v0, 0x3118 ; encoding: [0xff,0x02,0x00,0x7e,0x18,0x31,0x00,0x00]
-; GFX11-NEXT: s_mov_b32 s3, 0x31016000 ; encoding: [0xff,0x00,0x83,0xbe,0x00,0x60,0x01,0x31]
-; GFX11-NEXT: s_mov_b32 s2, -1 ; encoding: [0xc1,0x00,0x82,0xbe]
-; GFX11-NEXT: s_waitcnt lgkmcnt(0) ; encoding: [0x07,0xfc,0x89,0xbf]
-; GFX11-NEXT: buffer_store_b16 v0, off, s[0:3], 0 ; encoding: [0x00,0x00,0x64,0xe0,0x00,0x00,0x00,0x80]
-; GFX11-NEXT: s_endpgm ; encoding: [0x00,0x00,0xb0,0xbf]
+; GFX11-TRUE16-LABEL: store_inline_imm_inv_2pi_f16:
+; GFX11-TRUE16: ; %bb.0:
+; GFX11-TRUE16-NEXT: s_load_b64 s[0:1], s[4:5], 0x0 ; encoding: [0x02,0x00,0x04,0xf4,0x00,0x00,0x00,0xf8]
+; GFX11-TRUE16-NEXT: v_mov_b16_e32 v0.l, 0x3118 ; encoding: [0xff,0x38,0x00,0x7e,0x18,0x31,0x00,0x00]
+; GFX11-TRUE16-NEXT: s_mov_b32 s3, 0x31016000 ; encoding: [0xff,0x00,0x83,0xbe,0x00,0x60,0x01,0x31]
+; GFX11-TRUE16-NEXT: s_mov_b32 s2, -1 ; encoding: [0xc1,0x00,0x82,0xbe]
+; GFX11-TRUE16-NEXT: s_waitcnt lgkmcnt(0) ; encoding: [0x07,0xfc,0x89,0xbf]
+; GFX11-TRUE16-NEXT: buffer_store_b16 v0, off, s[0:3], 0 ; encoding: [0x00,0x00,0x64,0xe0,0x00,0x00,0x00,0x80]
+; GFX11-TRUE16-NEXT: s_endpgm ; encoding: [0x00,0x00,0xb0,0xbf]
+;
+; GFX11-FAKE16-LABEL: store_inline_imm_inv_2pi_f16:
+; GFX11-FAKE16: ; %bb.0:
+; GFX11-FAKE16-NEXT: s_load_b64 s[0:1], s[4:5], 0x0 ; encoding: [0x02,0x00,0x04,0xf4,0x00,0x00,0x00,0xf8]
+; GFX11-FAKE16-NEXT: v_mov_b32_e32 v0, 0x3118 ; encoding: [0xff,0x02,0x00,0x7e,0x18,0x31,0x00,0x00]
+; GFX11-FAKE16-NEXT: s_mov_b32 s3, 0x31016000 ; encoding: [0xff,0x00,0x83,0xbe,0x00,0x60,0x01,0x31]
+; GFX11-FAKE16-NEXT: s_mov_b32 s2, -1 ; encoding: [0xc1,0x00,0x82,0xbe]
+; GFX11-FAKE16-NEXT: s_waitcnt lgkmcnt(0) ; encoding: [0x07,0xfc,0x89,0xbf]
+; GFX11-FAKE16-NEXT: buffer_store_b16 v0, off, s[0:3], 0 ; encoding: [0x00,0x00,0x64,0xe0,0x00,0x00,0x00,0x80]
+; GFX11-FAKE16-NEXT: s_endpgm ; encoding: [0x00,0x00,0xb0,0xbf]
;
; VI-LABEL: store_inline_imm_inv_2pi_f16:
; VI: ; %bb.0:
@@ -550,15 +671,25 @@ define amdgpu_kernel void @store_inline_imm_m_inv_2pi_f16(ptr addrspace(1) %out)
; GFX10-NEXT: buffer_store_short v0, off, s[0:3], 0 ; encoding: [0x00,0x00,0x68,0xe0,0x00,0x00,0x00,0x80]
; GFX10-NEXT: s_endpgm ; encoding: [0x00,0x00,0x81,0xbf]
;
-; GFX11-LABEL: store_inline_imm_m_inv_2pi_f16:
-; GFX11: ; %bb.0:
-; GFX11-NEXT: s_load_b64 s[0:1], s[4:5], 0x0 ; encoding: [0x02,0x00,0x04,0xf4,0x00,0x00,0x00,0xf8]
-; GFX11-NEXT: v_mov_b32_e32 v0, 0xffffb118 ; encoding: [0xff,0x02,0x00,0x7e,0x18,0xb1,0xff,0xff]
-; GFX11-NEXT: s_mov_b32 s3, 0x31016000 ; encoding: [0xff,0x00,0x83,0xbe,0x00,0x60,0x01,0x31]
-; GFX11-NEXT: s_mov_b32 s2, -1 ; encoding: [0xc1,0x00,0x82,0xbe]
-; GFX11-NEXT: s_waitcnt lgkmcnt(0) ; encoding: [0x07,0xfc,0x89,0xbf]
-; GFX11-NEXT: buffer_store_b16 v0, off, s[0:3], 0 ; encoding: [0x00,0x00,0x64,0xe0,0x00,0x00,0x00,0x80]
-; GFX11-NEXT: s_endpgm ; encoding: [0x00,0x00,0xb0,0xbf]
+; GFX11-TRUE16-LABEL: store_inline_imm_m_inv_2pi_f16:
+; GFX11-TRUE16: ; %bb.0:
+; GFX11-TRUE16-NEXT: s_load_b64 s[0:1], s[4:5], 0x0 ; encoding: [0x02,0x00,0x04,0xf4,0x00,0x00,0x00,0xf8]
+; GFX11-TRUE16-NEXT: v_mov_b16_e32 v0.l, 0xb118 ; encoding: [0xff,0x38,0x00,0x7e,0x18,0xb1,0xff,0xff]
+; GFX11-TRUE16-NEXT: s_mov_b32 s3, 0x31016000 ; encoding: [0xff,0x00,0x83,0xbe,0x00,0x60,0x01,0x31]
+; GFX11-TRUE16-NEXT: s_mov_b32 s2, -1 ; encoding: [0xc1,0x00,0x82,0xbe]
+; GFX11-TRUE16-NEXT: s_waitcnt lgkmcnt(0) ; encoding: [0x07,0xfc,0x89,0xbf]
+; GFX11-TRUE16-NEXT: buffer_store_b16 v0, off, s[0:3], 0 ; encoding: [0x00,0x00,0x64,0xe0,0x00,0x00,0x00,0x80]
+; GFX11-TRUE16-NEXT: s_endpgm ; encoding: [0x00,0x00,0xb0,0xbf]
+;
+; GFX11-FAKE16-LABEL: store_inline_imm_m_inv_2pi_f16:
+; GFX11-FAKE16: ; %bb.0:
+; GFX11-FAKE16-NEXT: s_load_b64 s[0:1], s[4:5], 0x0 ; encoding: [0x02,0x00,0x04,0xf4,0x00,0x00,0x00,0xf8]
+; GFX11-FAKE16-NEXT: v_mov_b32_e32 v0, 0xffffb118 ; encoding: [0xff,0x02,0x00,0x7e,0x18,0xb1,0xff,0xff]
+; GFX11-FAKE16-NEXT: s_mov_b32 s3, 0x31016000 ; encoding: [0xff,0x00,0x83,0xbe,0x00,0x60,0x01,0x31]
+; GFX11-FAKE16-NEXT: s_mov_b32 s2, -1 ; encoding: [0xc1,0x00,0x82,0xbe]
+; GFX11-FAKE16-NEXT: s_waitcnt lgkmcnt(0) ; encoding: [0x07,0xfc,0x89,0xbf]
+; GFX11-FAKE16-NEXT: buffer_store_b16 v0, off, s[0:3], 0 ; encoding: [0x00,0x00,0x64,0xe0,0x00,0x00,0x00,0x80]
+; GFX11-FAKE16-NEXT: s_endpgm ; encoding: [0x00,0x00,0xb0,0xbf]
;
; VI-LABEL: store_inline_imm_m_inv_2pi_f16:
; VI: ; %bb.0:
@@ -594,15 +725,25 @@ define amdgpu_kernel void @store_literal_imm_f16(ptr addrspace(1) %out) {
; GFX10-NEXT: buffer_store_short v0, off, s[0:3], 0 ; encoding: [0x00,0x00,0x68,0xe0,0x00,0x00,0x00,0x80]
; GFX10-NEXT: s_endpgm ; encoding: [0x00,0x00,0x81,0xbf]
;
-; GFX11-LABEL: store_literal_imm_f16:
-; GFX11: ; %bb.0:
-; GFX11-NEXT: s_load_b64 s[0:1], s[4:5], 0x0 ; encoding: [0x02,0x00,0x04,0xf4,0x00,0x00,0x00,0xf8]
-; GFX11-NEXT: v_mov_b32_e32 v0, 0x6c00 ; encoding: [0xff,0x02,0x00,0x7e,0x00,0x6c,0x00,0x00]
-; GFX11-NEXT: s_mov_b32 s3, 0x31016000 ; encoding: [0xff,0x00,0x83,0xbe,0x00,0x60,0x01,0x31]
-; GFX11-NEXT: s_mov_b32 s2, -1 ; encoding: [0xc1,0x00,0x82,0xbe]
-; GFX11-NEXT: s_waitcnt lgkmcnt(0) ; encoding: [0x07,0xfc,0x89,0xbf]
-; GFX11-NEXT: buffer_store_b16 v0, off, s[0:3], 0 ; encoding: [0x00,0x00,0x64,0xe0,0x00,0x00,0x00,0x80]
-; GFX11-NEXT: s_endpgm ; encoding: [0x00,0x00,0xb0,0xbf]
+; GFX11-TRUE16-LABEL: store_literal_imm_f16:
+; GFX11-TRUE16: ; %bb.0:
+; GFX11-TRUE16-NEXT: s_load_b64 s[0:1], s[4:5], 0x0 ; encoding: [0x02,0x00,0x04,0xf4,0x00,0x00,0x00,0xf8]
+; GFX11-TRUE16-NEXT: v_mov_b16_e32 v0.l, 0x6c00 ; encoding: [0xff,0x38,0x00,0x7e,0x00,0x6c,0x00,0x00]
+; GFX11-TRUE16-NEXT: s_mov_b32 s3, 0x31016000 ; encoding: [0xff,0x00,0x83,0xbe,0x00,0x60,0x01,0x31]
+; GFX11-TRUE16-NEXT: s_mov_b32 s2, -1 ; encoding: [0xc1,0x00,0x82,0xbe]
+; GFX11-TRUE16-NEXT: s_waitcnt lgkmcnt(0) ; encoding: [0x07,0xfc,0x89,0xbf]
+; GFX11-TRUE16-NEXT: buffer_store_b16 v0, off, s[0:3], 0 ; encoding: [0x00,0x00,0x64,0xe0,0x00,0x00,0x00,0x80]
+; GFX11-TRUE16-NEXT: s_endpgm ; encoding: [0x00,0x00,0xb0,0xbf]
+;
+; GFX11-FAKE16-LABEL: store_literal_imm_f16:
+; GFX11-FAKE16: ; %bb.0:
+; GFX11-FAKE16-NEXT: s_load_b64 s[0:1], s[4:5], 0x0 ; encoding: [0x02,0x00,0x04,0xf4,0x00,0x00,0x00,0xf8]
+; GFX11-FAKE16-NEXT: v_mov_b32_e32 v0, 0x6c00 ; encoding: [0xff,0x02,0x00,0x7e,0x00,0x6c,0x00,0x00]
+; GFX11-FAKE16-NEXT: s_mov_b32 s3, 0x31016000 ; encoding: [0xff,0x00,0x83,0xbe,0x00,0x60,0x01,0x31]
+; GFX11-FAKE16-NEXT: s_mov_b32 s2, -1 ; encoding: [0xc1,0x00,0x82,0xbe]
+; GFX11-FAKE16-NEXT: s_waitcnt lgkmcnt(0) ; encoding: [0x07,0xfc,0x89,0xbf]
+; GFX11-FAKE16-NEXT: buffer_store_b16 v0, off, s[0:3], 0 ; encoding: [0x00,0x00,0x64,0xe0,0x00,0x00,0x00,0x80]
+; GFX11-FAKE16-NEXT: s_endpgm ; encoding: [0x00,0x00,0xb0,0xbf]
;
; VI-LABEL: store_literal_imm_f16:
; VI: ; %bb.0:
diff --git a/llvm/test/CodeGen/AMDGPU/llvm.amdgcn.interp.inreg.ll b/llvm/test/CodeGen/AMDGPU/llvm.amdgcn.interp.inreg.ll
index 46e2e9249ba95..d9226df0010f6 100644
--- a/llvm/test/CodeGen/AMDGPU/llvm.amdgcn.interp.inreg.ll
+++ b/llvm/test/CodeGen/AMDGPU/llvm.amdgcn.interp.inreg.ll
@@ -358,14 +358,15 @@ main_body:
define amdgpu_ps half @v_interp_f16_imm_params(float inreg %i, float inreg %j) #0 {
; GFX11-TRUE16-LABEL: v_interp_f16_imm_params:
; GFX11-TRUE16: ; %bb.0: ; %main_body
-; GFX11-TRUE16-NEXT: v_mov_b16_e32 v0.l, 0
-; GFX11-TRUE16-NEXT: v_dual_mov_b32 v1, s0 :: v_dual_mov_b32 v2, 0
+; GFX11-TRUE16-NEXT: v_dual_mov_b32 v1, 0 :: v_dual_mov_b32 v2, s0
; GFX11-TRUE16-NEXT: v_mov_b32_e32 v3, s1
-; GFX11-TRUE16-NEXT: s_delay_alu instid0(VALU_DEP_2) | instskip(NEXT) | instid1(VALU_DEP_2)
-; GFX11-TRUE16-NEXT: v_interp_p10_f16_f32 v1, v0.l, v1, v0.l wait_exp:7
-; GFX11-TRUE16-NEXT: v_interp_p2_f16_f32 v0.l, v0.l, v3, v2 wait_exp:7
; GFX11-TRUE16-NEXT: s_delay_alu instid0(VALU_DEP_2) | instskip(NEXT) | instid1(VALU_DEP_1)
-; GFX11-TRUE16-NEXT: v_cvt_f16_f32_e32 v0.h, v1
+; GFX11-TRUE16-NEXT: v_mov_b16_e32 v0.l, v1.l
+; GFX11-TRUE16-NEXT: v_interp_p10_f16_f32 v2, v0.l, v2, v0.l wait_exp:7
+; GFX11-TRUE16-NEXT: s_delay_alu instid0(VALU_DEP_3) | instskip(NEXT) | instid1(VALU_DEP_2)
+; GFX11-TRUE16-NEXT: v_interp_p2_f16_f32 v0.l, v0.l, v3, v1 wait_exp:7
+; GFX11-TRUE16-NEXT: v_cvt_f16_f32_e32 v0.h, v2
+; GFX11-TRUE16-NEXT: s_delay_alu instid0(VALU_DEP_1)
; GFX11-TRUE16-NEXT: v_add_f16_e32 v0.l, v0.h, v0.l
; GFX11-TRUE16-NEXT: ; return to shader part epilog
;
@@ -383,14 +384,15 @@ define amdgpu_ps half @v_interp_f16_imm_params(float inreg %i, float inreg %j) #
;
; GFX12-TRUE16-LABEL: v_interp_f16_imm_params:
; GFX12-TRUE16: ; %bb.0: ; %main_body
-; GFX12-TRUE16-NEXT: v_mov_b16_e32 v0.l, 0
-; GFX12-TRUE16-NEXT: v_dual_mov_b32 v1, s0 :: v_dual_mov_b32 v2, 0
+; GFX12-TRUE16-NEXT: v_dual_mov_b32 v1, 0 :: v_dual_mov_b32 v2, s0
; GFX12-TRUE16-NEXT: v_mov_b32_e32 v3, s1
-; GFX12-TRUE16-NEXT: s_delay_alu instid0(VALU_DEP_2) | instskip(NEXT) | instid1(VALU_DEP_2)
-; GFX12-TRUE16-NEXT: v_interp_p10_f16_f32 v1, v0.l, v1, v0.l wait_exp:7
-; GFX12-TRUE16-NEXT: v_interp_p2_f16_f32 v0.l, v0.l, v3, v2 wait_exp:7
; GFX12-TRUE16-NEXT: s_delay_alu instid0(VALU_DEP_2) | instskip(NEXT) | instid1(VALU_DEP_1)
-; GFX12-TRUE16-NEXT: v_cvt_f16_f32_e32 v0.h, v1
+; GFX12-TRUE16-NEXT: v_mov_b16_e32 v0.l, v1.l
+; GFX12-TRUE16-NEXT: v_interp_p10_f16_f32 v2, v0.l, v2, v0.l wait_exp:7
+; GFX12-TRUE16-NEXT: s_delay_alu instid0(VALU_DEP_3) | instskip(NEXT) | instid1(VALU_DEP_2)
+; GFX12-TRUE16-NEXT: v_interp_p2_f16_f32 v0.l, v0.l, v3, v1 wait_exp:7
+; GFX12-TRUE16-NEXT: v_cvt_f16_f32_e32 v0.h, v2
+; GFX12-TRUE16-NEXT: s_delay_alu instid0(VALU_DEP_1)
; GFX12-TRUE16-NEXT: v_add_f16_e32 v0.l, v0.h, v0.l
; GFX12-TRUE16-NEXT: ; return to shader part epilog
;
More information about the llvm-commits
mailing list