diff --git a/llvm/lib/Target/AMDGPU/SIInsertWaitcnts.cpp b/llvm/lib/Target/AMDGPU/SIInsertWaitcnts.cpp
index dd3f2fe25a23..520c321cb7ea 100644
--- a/llvm/lib/Target/AMDGPU/SIInsertWaitcnts.cpp
+++ b/llvm/lib/Target/AMDGPU/SIInsertWaitcnts.cpp
@@ -552,7 +552,7 @@ public:
         (!Inst.mayLoad() || SIInstrInfo::isAtomicNoRet(Inst))) {
       // FLAT and SCRATCH instructions may access scratch. Other VMEM
       // instructions do not.
-      if (SIInstrInfo::isFLAT(Inst) && mayAccessScratchThroughFlat(Inst))
+      if (TII->mayAccessScratchThroughFlat(Inst))
         return SCRATCH_WRITE_ACCESS;
       return VMEM_WRITE_ACCESS;
     }
@@ -565,7 +565,6 @@ public:
 
   bool mayAccessVMEMThroughFlat(const MachineInstr &MI) const;
   bool mayAccessLDSThroughFlat(const MachineInstr &MI) const;
-  bool mayAccessScratchThroughFlat(const MachineInstr &MI) const;
   bool isVmemAccess(const MachineInstr &MI) const;
   bool generateWaitcntInstBefore(MachineInstr &MI,
                                  WaitcntBrackets &ScoreBrackets,
@@ -2160,32 +2159,6 @@ bool SIInsertWaitcnts::mayAccessLDSThroughFlat(const MachineInstr &MI) const {
   return false;
 }
 
-// This is a flat memory operation. Check to see if it has memory tokens for
-// either scratch or FLAT.
-bool SIInsertWaitcnts::mayAccessScratchThroughFlat(
-    const MachineInstr &MI) const {
-  assert(TII->isFLAT(MI));
-
-  // SCRATCH instructions always access scratch.
-  if (TII->isFLATScratch(MI))
-    return true;
-
-  // GLOBAL instructions never access scratch.
-  if (TII->isFLATGlobal(MI))
-    return false;
-
-  // If there are no memory operands then conservatively assume the flat
-  // operation may access scratch.
-  if (MI.memoperands_empty())
-    return true;
-
-  // See if any memory operand specifies an address space that involves scratch.
-  return any_of(MI.memoperands(), [](const MachineMemOperand *Memop) {
-    unsigned AS = Memop->getAddrSpace();
-    return AS == AMDGPUAS::PRIVATE_ADDRESS || AS == AMDGPUAS::FLAT_ADDRESS;
-  });
-}
-
 bool SIInsertWaitcnts::isVmemAccess(const MachineInstr &MI) const {
   return (TII->isFLAT(MI) && mayAccessVMEMThroughFlat(MI)) ||
          (TII->isVMEM(MI) && !AMDGPU::getMUBUFIsBufferInv(MI.getOpcode()));
diff --git a/llvm/lib/Target/AMDGPU/SIInstrInfo.cpp b/llvm/lib/Target/AMDGPU/SIInstrInfo.cpp
index 8d6c1d046602..2aa6b4e82f9d 100644
--- a/llvm/lib/Target/AMDGPU/SIInstrInfo.cpp
+++ b/llvm/lib/Target/AMDGPU/SIInstrInfo.cpp
@@ -4249,6 +4249,32 @@ bool SIInstrInfo::isAlwaysGDS(uint16_t Opcode) const {
          Opcode == AMDGPU::DS_SUB_GS_REG_RTN || isGWS(Opcode);
 }
 
+bool SIInstrInfo::mayAccessScratchThroughFlat(const MachineInstr &MI) const {
+  if (!isFLAT(MI) || isFLATGlobal(MI))
+    return false;
+
+  // If scratch is not initialized, we can never access it.
+  if (MI.getMF()->getFunction().hasFnAttribute("amdgpu-no-flat-scratch-init"))
+    return false;
+
+  // SCRATCH instructions always access scratch.
+  if (isFLATScratch(MI))
+    return true;
+
+  // If there are no memory operands then conservatively assume the flat
+  // operation may access scratch.
+  if (MI.memoperands_empty())
+    return true;
+
+  // TODO (?): Does this need to be taught how to read noalias.addrspace ?
+
+  // See if any memory operand specifies an address space that involves scratch.
+  return any_of(MI.memoperands(), [](const MachineMemOperand *Memop) {
+    unsigned AS = Memop->getAddrSpace();
+    return AS == AMDGPUAS::PRIVATE_ADDRESS || AS == AMDGPUAS::FLAT_ADDRESS;
+  });
+}
+
 bool SIInstrInfo::modifiesModeRegister(const MachineInstr &MI) {
   // Skip the full operand and register alias search modifiesRegister
   // does. There's only a handful of instructions that touch this, it's only an
diff --git a/llvm/lib/Target/AMDGPU/SIInstrInfo.h b/llvm/lib/Target/AMDGPU/SIInstrInfo.h
index 2ffb7835c312..e042b59eb0f0 100644
--- a/llvm/lib/Target/AMDGPU/SIInstrInfo.h
+++ b/llvm/lib/Target/AMDGPU/SIInstrInfo.h
@@ -678,6 +678,12 @@ public:
     return get(Opcode).TSFlags & SIInstrFlags::FLAT;
   }
 
+  /// \returns true for SCRATCH_ instructions, or FLAT_ instructions with
+  /// SCRATCH_ memory operands.
+  /// Conservatively correct; will return true if \p MI cannot be proven
+  /// to not hit scratch.
+  bool mayAccessScratchThroughFlat(const MachineInstr &MI) const;
+
   static bool isBlockLoadStore(uint16_t Opcode) {
     switch (Opcode) {
     case AMDGPU::SI_BLOCK_SPILL_V1024_SAVE:
diff --git a/llvm/lib/Target/AMDGPU/SIMemoryLegalizer.cpp b/llvm/lib/Target/AMDGPU/SIMemoryLegalizer.cpp
index 0e8a420fbb70..607825ea50e7 100644
--- a/llvm/lib/Target/AMDGPU/SIMemoryLegalizer.cpp
+++ b/llvm/lib/Target/AMDGPU/SIMemoryLegalizer.cpp
@@ -321,7 +321,8 @@ public:
                                               bool IsNonTemporal,
                                               bool IsLastUse = false) const = 0;
 
-  virtual bool expandSystemScopeStore(MachineBasicBlock::iterator &MI) const {
+  virtual bool finalizeStore(MachineBasicBlock::iterator &MI,
+                             bool Atomic) const {
     return false;
   };
 
@@ -602,7 +603,8 @@ public:
                                       bool IsVolatile, bool IsNonTemporal,
                                       bool IsLastUse) const override;
 
-  bool expandSystemScopeStore(MachineBasicBlock::iterator &MI) const override;
+  bool finalizeStore(MachineBasicBlock::iterator &MI,
+                     bool Atomic) const override;
 
   bool insertRelease(MachineBasicBlock::iterator &MI, SIAtomicScope Scope,
                      SIAtomicAddrSpace AddrSpace, bool IsCrossAddrSpaceOrdering,
@@ -2551,11 +2553,25 @@ bool SIGfx12CacheControl::enableVolatileAndOrNonTemporal(
   return Changed;
 }
 
-bool SIGfx12CacheControl::expandSystemScopeStore(
-    MachineBasicBlock::iterator &MI) const {
+bool SIGfx12CacheControl::finalizeStore(MachineBasicBlock::iterator &MI,
+                                        bool Atomic) const {
   MachineOperand *CPol = TII->getNamedOperand(*MI, OpName::cpol);
-  if (CPol && ((CPol->getImm() & CPol::SCOPE) == CPol::SCOPE_SYS))
-    return insertWaitsBeforeSystemScopeStore(MI);
+  if (!CPol)
+    return false;
+
+  const unsigned Scope = CPol->getImm() & CPol::SCOPE;
+
+  // GFX12.0 only: Extra waits needed before system scope stores.
+  if (!ST.hasGFX1250Insts()) {
+    if (!Atomic && Scope == CPol::SCOPE_SYS)
+      return insertWaitsBeforeSystemScopeStore(MI);
+    return false;
+  }
+
+  // GFX12.5 only: Require SCOPE_SE on stores that may hit the scratch address
+  // space.
+  if (TII->mayAccessScratchThroughFlat(*MI) && Scope == CPol::SCOPE_CU)
+    return setScope(MI, CPol::SCOPE_SE);
 
   return false;
 }
@@ -2674,6 +2690,7 @@ bool SIMemoryLegalizer::expandStore(const SIMemOpInfo &MOI,
                                    MOI.getIsCrossAddressSpaceOrdering(),
                                    Position::BEFORE);
 
+    Changed |= CC->finalizeStore(MI, /*Atomic=*/true);
     return Changed;
   }
 
@@ -2686,7 +2703,7 @@ bool SIMemoryLegalizer::expandStore(const SIMemOpInfo &MOI,
 
   // GFX12 specific, scope(desired coherence domain in cache hierarchy) is
   // instruction field, do not confuse it with atomic scope.
-  Changed |= CC->expandSystemScopeStore(MI);
+  Changed |= CC->finalizeStore(MI, /*Atomic=*/false);
   return Changed;
 }
 
diff --git a/llvm/test/CodeGen/AMDGPU/gfx1250-scratch-scope-se.ll b/llvm/test/CodeGen/AMDGPU/gfx1250-scratch-scope-se.ll
new file mode 100644
index 000000000000..d1e82a06077f
--- /dev/null
+++ b/llvm/test/CodeGen/AMDGPU/gfx1250-scratch-scope-se.ll
@@ -0,0 +1,95 @@
+; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py UTC_ARGS: --version 5
+; RUN: llc -mtriple=amdgcn-amd-amdhsa -mcpu=gfx1250 < %s | FileCheck -check-prefixes=GCN,GCN-SDAG %s
+; RUN: llc -global-isel -mtriple=amdgcn-amd-amdhsa -mcpu=gfx1250 < %s | FileCheck -check-prefixes=GCN,GCN-GISEL %s
+
+; Test that stores that may hit scratch are correctly promoted to SCOPE_SE.
+
+define void @test_scratch_store(ptr addrspace(5) %ptr, i32 %val) {
+; GCN-LABEL: test_scratch_store:
+; GCN:       ; %bb.0:
+; GCN-NEXT:    s_wait_loadcnt_dscnt 0x0
+; GCN-NEXT:    s_wait_kmcnt 0x0
+; GCN-NEXT:    scratch_store_b32 v0, v1, off scope:SCOPE_SE
+; GCN-NEXT:    s_set_pc_i64 s[30:31]
+    store i32 %val, ptr addrspace(5) %ptr
+    ret void
+}
+
+define void @test_unknown_flat_store(ptr %ptr, i32 %val) {
+; GCN-LABEL: test_unknown_flat_store:
+; GCN:       ; %bb.0:
+; GCN-NEXT:    s_wait_loadcnt_dscnt 0x0
+; GCN-NEXT:    s_wait_kmcnt 0x0
+; GCN-NEXT:    flat_store_b32 v[0:1], v2 scope:SCOPE_SE
+; GCN-NEXT:    s_wait_dscnt 0x0
+; GCN-NEXT:    s_set_pc_i64 s[30:31]
+    store i32 %val, ptr %ptr
+    ret void
+}
+
+define void @test_flat_store_no_scratch_alloc(ptr %ptr, i32 %val) #0 {
+; GCN-LABEL: test_flat_store_no_scratch_alloc:
+; GCN:       ; %bb.0:
+; GCN-NEXT:    s_wait_loadcnt_dscnt 0x0
+; GCN-NEXT:    s_wait_kmcnt 0x0
+; GCN-NEXT:    flat_store_b32 v[0:1], v2
+; GCN-NEXT:    s_wait_dscnt 0x0
+; GCN-NEXT:    s_set_pc_i64 s[30:31]
+    store i32 %val, ptr %ptr
+    ret void
+}
+
+; TODO: handle
+define void @test_flat_store_noalias_addrspace(ptr %ptr, i32 %val) {
+; GCN-LABEL: test_flat_store_noalias_addrspace:
+; GCN:       ; %bb.0:
+; GCN-NEXT:    s_wait_loadcnt_dscnt 0x0
+; GCN-NEXT:    s_wait_kmcnt 0x0
+; GCN-NEXT:    flat_store_b32 v[0:1], v2 scope:SCOPE_SE
+; GCN-NEXT:    s_wait_dscnt 0x0
+; GCN-NEXT:    s_set_pc_i64 s[30:31]
+    store i32 %val, ptr %ptr, !noalias.addrspace !{i32 5, i32 6}
+    ret void
+}
+
+; TODO: would be nice to handle too
+define void @test_flat_store_select(ptr addrspace(1) %a, ptr addrspace(3) %b, i1 %cond, i32 %val) {
+; GCN-SDAG-LABEL: test_flat_store_select:
+; GCN-SDAG:       ; %bb.0:
+; GCN-SDAG-NEXT:    s_wait_loadcnt_dscnt 0x0
+; GCN-SDAG-NEXT:    s_wait_kmcnt 0x0
+; GCN-SDAG-NEXT:    v_cmp_ne_u32_e32 vcc_lo, -1, v2
+; GCN-SDAG-NEXT:    v_and_b32_e32 v3, 1, v3
+; GCN-SDAG-NEXT:    s_mov_b64 s[0:1], src_shared_base
+; GCN-SDAG-NEXT:    v_cndmask_b32_e32 v2, 0, v2, vcc_lo
+; GCN-SDAG-NEXT:    v_cndmask_b32_e64 v5, 0, s1, vcc_lo
+; GCN-SDAG-NEXT:    s_delay_alu instid0(VALU_DEP_3) | instskip(NEXT) | instid1(VALU_DEP_2)
+; GCN-SDAG-NEXT:    v_cmp_eq_u32_e32 vcc_lo, 1, v3
+; GCN-SDAG-NEXT:    v_dual_cndmask_b32 v1, v5, v1 :: v_dual_cndmask_b32 v0, v2, v0
+; GCN-SDAG-NEXT:    flat_store_b32 v[0:1], v4 scope:SCOPE_SE
+; GCN-SDAG-NEXT:    s_wait_dscnt 0x0
+; GCN-SDAG-NEXT:    s_set_pc_i64 s[30:31]
+;
+; GCN-GISEL-LABEL: test_flat_store_select:
+; GCN-GISEL:       ; %bb.0:
+; GCN-GISEL-NEXT:    s_wait_loadcnt_dscnt 0x0
+; GCN-GISEL-NEXT:    s_wait_kmcnt 0x0
+; GCN-GISEL-NEXT:    v_cmp_ne_u32_e32 vcc_lo, -1, v2
+; GCN-GISEL-NEXT:    v_and_b32_e32 v3, 1, v3
+; GCN-GISEL-NEXT:    s_mov_b64 s[0:1], src_shared_base
+; GCN-GISEL-NEXT:    v_cndmask_b32_e32 v2, 0, v2, vcc_lo
+; GCN-GISEL-NEXT:    v_cndmask_b32_e64 v5, 0, s1, vcc_lo
+; GCN-GISEL-NEXT:    s_delay_alu instid0(VALU_DEP_3) | instskip(NEXT) | instid1(VALU_DEP_2)
+; GCN-GISEL-NEXT:    v_cmp_ne_u32_e32 vcc_lo, 0, v3
+; GCN-GISEL-NEXT:    v_dual_cndmask_b32 v0, v2, v0 :: v_dual_cndmask_b32 v1, v5, v1
+; GCN-GISEL-NEXT:    flat_store_b32 v[0:1], v4 scope:SCOPE_SE
+; GCN-GISEL-NEXT:    s_wait_dscnt 0x0
+; GCN-GISEL-NEXT:    s_set_pc_i64 s[30:31]
+    %a.ascast = addrspacecast ptr addrspace(1) %a to ptr
+    %b.ascast = addrspacecast ptr addrspace(3) %b to ptr
+    %ptr = select i1 %cond, ptr %a.ascast, ptr %b.ascast
+    store i32 %val, ptr %ptr
+    ret void
+}
+
+attributes #0 = { "amdgpu-no-flat-scratch-init" }
diff --git a/llvm/test/CodeGen/AMDGPU/global-load-xcnt.ll b/llvm/test/CodeGen/AMDGPU/global-load-xcnt.ll
index fd644a35f61e..3a898a921446 100644
--- a/llvm/test/CodeGen/AMDGPU/global-load-xcnt.ll
+++ b/llvm/test/CodeGen/AMDGPU/global-load-xcnt.ll
@@ -124,27 +124,27 @@ define i32 @test_v64i32_load_store(ptr addrspace(1) %ptr, i32 %idx, ptr addrspac
 ; GCN-SDAG-NEXT:    s_wait_loadcnt_dscnt 0x0
 ; GCN-SDAG-NEXT:    s_wait_kmcnt 0x0
 ; GCN-SDAG-NEXT:    s_clause 0xd
-; GCN-SDAG-NEXT:    scratch_store_b32 off, v40, s32 offset:52
-; GCN-SDAG-NEXT:    scratch_store_b32 off, v41, s32 offset:48
-; GCN-SDAG-NEXT:    scratch_store_b32 off, v42, s32 offset:44
-; GCN-SDAG-NEXT:    scratch_store_b32 off, v43, s32 offset:40
-; GCN-SDAG-NEXT:    scratch_store_b32 off, v44, s32 offset:36
-; GCN-SDAG-NEXT:    scratch_store_b32 off, v45, s32 offset:32
-; GCN-SDAG-NEXT:    scratch_store_b32 off, v56, s32 offset:28
-; GCN-SDAG-NEXT:    scratch_store_b32 off, v57, s32 offset:24
-; GCN-SDAG-NEXT:    scratch_store_b32 off, v58, s32 offset:20
-; GCN-SDAG-NEXT:    scratch_store_b32 off, v59, s32 offset:16
-; GCN-SDAG-NEXT:    scratch_store_b32 off, v60, s32 offset:12
-; GCN-SDAG-NEXT:    scratch_store_b32 off, v61, s32 offset:8
-; GCN-SDAG-NEXT:    scratch_store_b32 off, v62, s32 offset:4
-; GCN-SDAG-NEXT:    scratch_store_b32 off, v63, s32
+; GCN-SDAG-NEXT:    scratch_store_b32 off, v40, s32 offset:52 scope:SCOPE_SE
+; GCN-SDAG-NEXT:    scratch_store_b32 off, v41, s32 offset:48 scope:SCOPE_SE
+; GCN-SDAG-NEXT:    scratch_store_b32 off, v42, s32 offset:44 scope:SCOPE_SE
+; GCN-SDAG-NEXT:    scratch_store_b32 off, v43, s32 offset:40 scope:SCOPE_SE
+; GCN-SDAG-NEXT:    scratch_store_b32 off, v44, s32 offset:36 scope:SCOPE_SE
+; GCN-SDAG-NEXT:    scratch_store_b32 off, v45, s32 offset:32 scope:SCOPE_SE
+; GCN-SDAG-NEXT:    scratch_store_b32 off, v56, s32 offset:28 scope:SCOPE_SE
+; GCN-SDAG-NEXT:    scratch_store_b32 off, v57, s32 offset:24 scope:SCOPE_SE
+; GCN-SDAG-NEXT:    scratch_store_b32 off, v58, s32 offset:20 scope:SCOPE_SE
+; GCN-SDAG-NEXT:    scratch_store_b32 off, v59, s32 offset:16 scope:SCOPE_SE
+; GCN-SDAG-NEXT:    scratch_store_b32 off, v60, s32 offset:12 scope:SCOPE_SE
+; GCN-SDAG-NEXT:    scratch_store_b32 off, v61, s32 offset:8 scope:SCOPE_SE
+; GCN-SDAG-NEXT:    scratch_store_b32 off, v62, s32 offset:4 scope:SCOPE_SE
+; GCN-SDAG-NEXT:    scratch_store_b32 off, v63, s32 scope:SCOPE_SE
 ; GCN-SDAG-NEXT:    global_load_b128 v[6:9], v[0:1], off offset:224
 ; GCN-SDAG-NEXT:    v_dual_mov_b32 v5, v4 :: v_dual_mov_b32 v4, v3
 ; GCN-SDAG-NEXT:    s_wait_loadcnt 0x0
-; GCN-SDAG-NEXT:    scratch_store_b128 off, v[6:9], s32 offset:56 ; 16-byte Folded Spill
+; GCN-SDAG-NEXT:    scratch_store_b128 off, v[6:9], s32 offset:56 scope:SCOPE_SE ; 16-byte Folded Spill
 ; GCN-SDAG-NEXT:    global_load_b128 v[6:9], v[0:1], off offset:240
 ; GCN-SDAG-NEXT:    s_wait_loadcnt 0x0
-; GCN-SDAG-NEXT:    scratch_store_b128 off, v[6:9], s32 offset:72 ; 16-byte Folded Spill
+; GCN-SDAG-NEXT:    scratch_store_b128 off, v[6:9], s32 offset:72 scope:SCOPE_SE ; 16-byte Folded Spill
 ; GCN-SDAG-NEXT:    s_clause 0xd
 ; GCN-SDAG-NEXT:    global_load_b128 v[10:13], v[0:1], off offset:192
 ; GCN-SDAG-NEXT:    global_load_b128 v[14:17], v[0:1], off offset:208
@@ -206,27 +206,27 @@ define i32 @test_v64i32_load_store(ptr addrspace(1) %ptr, i32 %idx, ptr addrspac
 ; GCN-GISEL-NEXT:    s_wait_loadcnt_dscnt 0x0
 ; GCN-GISEL-NEXT:    s_wait_kmcnt 0x0
 ; GCN-GISEL-NEXT:    s_clause 0xf
-; GCN-GISEL-NEXT:    scratch_store_b32 off, v40, s32 offset:60
-; GCN-GISEL-NEXT:    scratch_store_b32 off, v41, s32 offset:56
-; GCN-GISEL-NEXT:    scratch_store_b32 off, v42, s32 offset:52
-; GCN-GISEL-NEXT:    scratch_store_b32 off, v43, s32 offset:48
-; GCN-GISEL-NEXT:    scratch_store_b32 off, v44, s32 offset:44
-; GCN-GISEL-NEXT:    scratch_store_b32 off, v45, s32 offset:40
-; GCN-GISEL-NEXT:    scratch_store_b32 off, v46, s32 offset:36
-; GCN-GISEL-NEXT:    scratch_store_b32 off, v47, s32 offset:32
-; GCN-GISEL-NEXT:    scratch_store_b32 off, v56, s32 offset:28
-; GCN-GISEL-NEXT:    scratch_store_b32 off, v57, s32 offset:24
-; GCN-GISEL-NEXT:    scratch_store_b32 off, v58, s32 offset:20
-; GCN-GISEL-NEXT:    scratch_store_b32 off, v59, s32 offset:16
-; GCN-GISEL-NEXT:    scratch_store_b32 off, v60, s32 offset:12
-; GCN-GISEL-NEXT:    scratch_store_b32 off, v61, s32 offset:8
-; GCN-GISEL-NEXT:    scratch_store_b32 off, v62, s32 offset:4
-; GCN-GISEL-NEXT:    scratch_store_b32 off, v63, s32
+; GCN-GISEL-NEXT:    scratch_store_b32 off, v40, s32 offset:60 scope:SCOPE_SE
+; GCN-GISEL-NEXT:    scratch_store_b32 off, v41, s32 offset:56 scope:SCOPE_SE
+; GCN-GISEL-NEXT:    scratch_store_b32 off, v42, s32 offset:52 scope:SCOPE_SE
+; GCN-GISEL-NEXT:    scratch_store_b32 off, v43, s32 offset:48 scope:SCOPE_SE
+; GCN-GISEL-NEXT:    scratch_store_b32 off, v44, s32 offset:44 scope:SCOPE_SE
+; GCN-GISEL-NEXT:    scratch_store_b32 off, v45, s32 offset:40 scope:SCOPE_SE
+; GCN-GISEL-NEXT:    scratch_store_b32 off, v46, s32 offset:36 scope:SCOPE_SE
+; GCN-GISEL-NEXT:    scratch_store_b32 off, v47, s32 offset:32 scope:SCOPE_SE
+; GCN-GISEL-NEXT:    scratch_store_b32 off, v56, s32 offset:28 scope:SCOPE_SE
+; GCN-GISEL-NEXT:    scratch_store_b32 off, v57, s32 offset:24 scope:SCOPE_SE
+; GCN-GISEL-NEXT:    scratch_store_b32 off, v58, s32 offset:20 scope:SCOPE_SE
+; GCN-GISEL-NEXT:    scratch_store_b32 off, v59, s32 offset:16 scope:SCOPE_SE
+; GCN-GISEL-NEXT:    scratch_store_b32 off, v60, s32 offset:12 scope:SCOPE_SE
+; GCN-GISEL-NEXT:    scratch_store_b32 off, v61, s32 offset:8 scope:SCOPE_SE
+; GCN-GISEL-NEXT:    scratch_store_b32 off, v62, s32 offset:4 scope:SCOPE_SE
+; GCN-GISEL-NEXT:    scratch_store_b32 off, v63, s32 scope:SCOPE_SE
 ; GCN-GISEL-NEXT:    s_wait_xcnt 0x8
 ; GCN-GISEL-NEXT:    v_dual_mov_b32 v46, v3 :: v_dual_mov_b32 v47, v4
 ; GCN-GISEL-NEXT:    global_load_b128 v[2:5], v[0:1], off offset:32
 ; GCN-GISEL-NEXT:    s_wait_loadcnt 0x0
-; GCN-GISEL-NEXT:    scratch_store_b128 off, v[2:5], s32 offset:80 ; 16-byte Folded Spill
+; GCN-GISEL-NEXT:    scratch_store_b128 off, v[2:5], s32 offset:80 scope:SCOPE_SE ; 16-byte Folded Spill
 ; GCN-GISEL-NEXT:    s_clause 0xe
 ; GCN-GISEL-NEXT:    global_load_b128 v[6:9], v[0:1], off offset:48
 ; GCN-GISEL-NEXT:    global_load_b128 v[10:13], v[0:1], off offset:64
@@ -244,7 +244,7 @@ define i32 @test_v64i32_load_store(ptr addrspace(1) %ptr, i32 %idx, ptr addrspac
 ; GCN-GISEL-NEXT:    global_load_b128 v[60:63], v[0:1], off offset:16
 ; GCN-GISEL-NEXT:    global_load_b128 v[0:3], v[0:1], off offset:240
 ; GCN-GISEL-NEXT:    s_wait_loadcnt 0x0
-; GCN-GISEL-NEXT:    scratch_store_b128 off, v[0:3], s32 offset:64 ; 16-byte Folded Spill
+; GCN-GISEL-NEXT:    scratch_store_b128 off, v[0:3], s32 offset:64 scope:SCOPE_SE ; 16-byte Folded Spill
 ; GCN-GISEL-NEXT:    scratch_load_b128 v[0:3], off, s32 offset:80 th:TH_LOAD_LU ; 16-byte Folded Reload
 ; GCN-GISEL-NEXT:    s_wait_loadcnt 0x0
 ; GCN-GISEL-NEXT:    s_clause 0xe
@@ -299,10 +299,10 @@ define i64 @test_v16i64_load_store(ptr addrspace(1) %ptr_a, ptr addrspace(1) %pt
 ; GCN-SDAG-NEXT:    s_wait_loadcnt_dscnt 0x0
 ; GCN-SDAG-NEXT:    s_wait_kmcnt 0x0
 ; GCN-SDAG-NEXT:    s_clause 0x3
-; GCN-SDAG-NEXT:    scratch_store_b32 off, v40, s32 offset:12
-; GCN-SDAG-NEXT:    scratch_store_b32 off, v41, s32 offset:8
-; GCN-SDAG-NEXT:    scratch_store_b32 off, v42, s32 offset:4
-; GCN-SDAG-NEXT:    scratch_store_b32 off, v43, s32
+; GCN-SDAG-NEXT:    scratch_store_b32 off, v40, s32 offset:12 scope:SCOPE_SE
+; GCN-SDAG-NEXT:    scratch_store_b32 off, v41, s32 offset:8 scope:SCOPE_SE
+; GCN-SDAG-NEXT:    scratch_store_b32 off, v42, s32 offset:4 scope:SCOPE_SE
+; GCN-SDAG-NEXT:    scratch_store_b32 off, v43, s32 scope:SCOPE_SE
 ; GCN-SDAG-NEXT:    s_clause 0x7
 ; GCN-SDAG-NEXT:    global_load_b128 v[10:13], v[0:1], off offset:112
 ; GCN-SDAG-NEXT:    global_load_b128 v[18:21], v[0:1], off offset:96
@@ -385,12 +385,12 @@ define i64 @test_v16i64_load_store(ptr addrspace(1) %ptr_a, ptr addrspace(1) %pt
 ; GCN-GISEL-NEXT:    s_wait_loadcnt_dscnt 0x0
 ; GCN-GISEL-NEXT:    s_wait_kmcnt 0x0
 ; GCN-GISEL-NEXT:    s_clause 0x5
-; GCN-GISEL-NEXT:    scratch_store_b32 off, v40, s32 offset:20
-; GCN-GISEL-NEXT:    scratch_store_b32 off, v41, s32 offset:16
-; GCN-GISEL-NEXT:    scratch_store_b32 off, v42, s32 offset:12
-; GCN-GISEL-NEXT:    scratch_store_b32 off, v43, s32 offset:8
-; GCN-GISEL-NEXT:    scratch_store_b32 off, v44, s32 offset:4
-; GCN-GISEL-NEXT:    scratch_store_b32 off, v45, s32
+; GCN-GISEL-NEXT:    scratch_store_b32 off, v40, s32 offset:20 scope:SCOPE_SE
+; GCN-GISEL-NEXT:    scratch_store_b32 off, v41, s32 offset:16 scope:SCOPE_SE
+; GCN-GISEL-NEXT:    scratch_store_b32 off, v42, s32 offset:12 scope:SCOPE_SE
+; GCN-GISEL-NEXT:    scratch_store_b32 off, v43, s32 offset:8 scope:SCOPE_SE
+; GCN-GISEL-NEXT:    scratch_store_b32 off, v44, s32 offset:4 scope:SCOPE_SE
+; GCN-GISEL-NEXT:    scratch_store_b32 off, v45, s32 scope:SCOPE_SE
 ; GCN-GISEL-NEXT:    s_clause 0x7
 ; GCN-GISEL-NEXT:    global_load_b128 v[6:9], v[0:1], off offset:80
 ; GCN-GISEL-NEXT:    global_load_b128 v[10:13], v[0:1], off