diff --git a/llvm/lib/Target/RISCV/RISCVFrameLowering.cpp b/llvm/lib/Target/RISCV/RISCVFrameLowering.cpp
index 22824b77c37dd..44d73c3ca57f1 100644
--- a/llvm/lib/Target/RISCV/RISCVFrameLowering.cpp
+++ b/llvm/lib/Target/RISCV/RISCVFrameLowering.cpp
@@ -168,7 +168,8 @@ static void emitSCSEpilogue(MachineFunction &MF, MachineBasicBlock &MBB,
 // restored by the libcall, except it is zero-indexed - ID 0 corresponds to a
 // single register.
 static int getLibCallID(const MachineFunction &MF,
-                        const std::vector<CalleeSavedInfo> &CSI) {
+                        const std::vector<CalleeSavedInfo> &CSI,
+                        RISCVABI::ABI ABI) {
   const auto *RVFI = MF.getInfo<RISCVMachineFunctionInfo>();
 
   if (CSI.empty() || !RVFI->useSaveRestoreLibCalls(MF))
@@ -184,6 +185,14 @@ static int getLibCallID(const MachineFunction &MF,
   if (MaxReg == RISCV::NoRegister)
     return -1;
 
+  // The libcall always save/restore ra/s0/s1 with the ABI ilp32e and lp64e.
+  if (ABI == RISCVABI::ABI_ILP32E || ABI == RISCVABI::ABI_LP64E) {
+    assert(
+        (MaxReg == RISCV::X1 || MaxReg == RISCV::X8 || MaxReg == RISCV::X9) &&
+        "Invalid MaxReg for ABI ilp32e and lp64e");
+    MaxReg = RISCV::X9;
+  }
+
   switch (MaxReg) {
   default:
     llvm_unreachable("Something has gone wrong!");
@@ -205,9 +214,9 @@ static int getLibCallID(const MachineFunction &MF,
 
 // Get the name of the libcall used for spilling callee saved registers.
 // If this function will not use save/restore libcalls, then return a nullptr.
-static const char *
-getSpillLibCallName(const MachineFunction &MF,
-                    const std::vector<CalleeSavedInfo> &CSI) {
+static const char *getSpillLibCallName(const MachineFunction &MF,
+                                       const std::vector<CalleeSavedInfo> &CSI,
+                                       RISCVABI::ABI ABI) {
   static const char *const SpillLibCalls[] = {
     "__riscv_save_0",
     "__riscv_save_1",
@@ -224,7 +233,7 @@ getSpillLibCallName(const MachineFunction &MF,
     "__riscv_save_12"
   };
 
-  int LibCallID = getLibCallID(MF, CSI);
+  int LibCallID = getLibCallID(MF, CSI, ABI);
   if (LibCallID == -1)
     return nullptr;
   return SpillLibCalls[LibCallID];
@@ -234,7 +243,8 @@ getSpillLibCallName(const MachineFunction &MF,
 // If this function will not use save/restore libcalls, then return a nullptr.
 static const char *
 getRestoreLibCallName(const MachineFunction &MF,
-                      const std::vector<CalleeSavedInfo> &CSI) {
+                      const std::vector<CalleeSavedInfo> &CSI,
+                      RISCVABI::ABI ABI) {
   static const char *const RestoreLibCalls[] = {
     "__riscv_restore_0",
     "__riscv_restore_1",
@@ -251,7 +261,7 @@ getRestoreLibCallName(const MachineFunction &MF,
     "__riscv_restore_12"
   };
 
-  int LibCallID = getLibCallID(MF, CSI);
+  int LibCallID = getLibCallID(MF, CSI, ABI);
   if (LibCallID == -1)
     return nullptr;
   return RestoreLibCalls[LibCallID];
@@ -573,8 +583,9 @@ void RISCVFrameLowering::emitPrologue(MachineFunction &MF,
   // For negative frame indices, the offset from the frame pointer will differ
   // depending on which of these groups the frame index applies to.
   // The following calculates the correct offset knowing the number of callee
-  // saved registers spilt by the two methods.
-  if (int LibCallRegs = getLibCallID(MF, MFI.getCalleeSavedInfo()) + 1) {
+  // saved registers spilt by the two methods.STI.getTargetABI()
+  if (int LibCallRegs =
+          getLibCallID(MF, MFI.getCalleeSavedInfo(), STI.getTargetABI()) + 1) {
     // Calculate the size of the frame managed by the libcall. The stack
     // alignment of these libcalls should be the same as how we set it in
     // getABIStackAlignment.
@@ -1463,7 +1474,7 @@ bool RISCVFrameLowering::assignCalleeSavedSpillSlots(
   if (RVFI->isPushable(MF)) {
     if (int64_t PushSize = RVFI->getRVPushStackSize())
       MFI.CreateFixedSpillStackObject(PushSize, -PushSize);
-  } else if (int LibCallRegs = getLibCallID(MF, CSI) + 1) {
+  } else if (int LibCallRegs = getLibCallID(MF, CSI, STI.getTargetABI()) + 1) {
     int64_t LibCallFrameSize =
         alignTo((STI.getXLen() / 8) * LibCallRegs, getStackAlign());
     MFI.CreateFixedSpillStackObject(LibCallFrameSize, -LibCallFrameSize);
@@ -1500,7 +1511,8 @@ bool RISCVFrameLowering::spillCalleeSavedRegisters(
       for (unsigned i = 0; i < PushedRegNum; i++)
         PushBuilder.addUse(FixedCSRFIMap[i].first, RegState::Implicit);
     }
-  } else if (const char *SpillLibCall = getSpillLibCallName(*MF, CSI)) {
+  } else if (const char *SpillLibCall =
+                 getSpillLibCallName(*MF, CSI, STI.getTargetABI())) {
     // Add spill libcall via non-callee-saved register t0.
     BuildMI(MBB, MI, DL, TII.get(RISCV::PseudoCALLReg), RISCV::X5)
         .addExternalSymbol(SpillLibCall, RISCVII::MO_CALL)
@@ -1625,7 +1637,8 @@ bool RISCVFrameLowering::restoreCalleeSavedRegisters(
         PopBuilder.addDef(FixedCSRFIMap[i].first, RegState::ImplicitDefine);
     }
   } else {
-    const char *RestoreLibCall = getRestoreLibCallName(*MF, CSI);
+    const char *RestoreLibCall =
+        getRestoreLibCallName(*MF, CSI, STI.getTargetABI());
     if (RestoreLibCall) {
       // Add restore libcall via tail call.
       MachineBasicBlock::iterator NewMI =
diff --git a/llvm/test/CodeGen/RISCV/calling-conv-ilp32e.ll b/llvm/test/CodeGen/RISCV/calling-conv-ilp32e.ll
index d08cf577b1bdd..e7c3e91702491 100644
--- a/llvm/test/CodeGen/RISCV/calling-conv-ilp32e.ll
+++ b/llvm/test/CodeGen/RISCV/calling-conv-ilp32e.ll
@@ -54,15 +54,15 @@ define i32 @callee_float_in_regs(i32 %a, float %b) {
 ;
 ; ILP32E-FPELIM-SAVE-RESTORE-LABEL: callee_float_in_regs:
 ; ILP32E-FPELIM-SAVE-RESTORE:       # %bb.0:
-; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    call t0, __riscv_save_1
-; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    .cfi_def_cfa_offset 8
+; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    call t0, __riscv_save_2
+; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    .cfi_def_cfa_offset 12
 ; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    .cfi_offset ra, -4
 ; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    .cfi_offset s0, -8
 ; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    mv s0, a0
 ; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    mv a0, a1
 ; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    call __fixsfsi
 ; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    add a0, s0, a0
-; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    tail __riscv_restore_1
+; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    tail __riscv_restore_2
 ;
 ; ILP32E-WITHFP-SAVE-RESTORE-LABEL: callee_float_in_regs:
 ; ILP32E-WITHFP-SAVE-RESTORE:       # %bb.0:
@@ -117,26 +117,26 @@ define i32 @caller_float_in_regs() {
 ;
 ; ILP32E-FPELIM-SAVE-RESTORE-LABEL: caller_float_in_regs:
 ; ILP32E-FPELIM-SAVE-RESTORE:       # %bb.0:
-; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    call t0, __riscv_save_0
-; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    .cfi_def_cfa_offset 4
+; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    call t0, __riscv_save_2
+; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    .cfi_def_cfa_offset 12
 ; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    .cfi_offset ra, -4
 ; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    li a0, 1
 ; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    lui a1, 262144
 ; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    call callee_float_in_regs
-; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    tail __riscv_restore_0
+; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    tail __riscv_restore_2
 ;
 ; ILP32E-WITHFP-SAVE-RESTORE-LABEL: caller_float_in_regs:
 ; ILP32E-WITHFP-SAVE-RESTORE:       # %bb.0:
-; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    call t0, __riscv_save_1
-; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    .cfi_def_cfa_offset 8
+; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    call t0, __riscv_save_2
+; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    .cfi_def_cfa_offset 12
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    .cfi_offset ra, -4
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    .cfi_offset s0, -8
-; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    addi s0, sp, 8
+; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    addi s0, sp, 12
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    .cfi_def_cfa s0, 0
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    li a0, 1
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    lui a1, 262144
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    call callee_float_in_regs
-; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    tail __riscv_restore_1
+; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    tail __riscv_restore_2
   %1 = call i32 @callee_float_in_regs(i32 1, float 2.0)
   ret i32 %1
 }
@@ -176,16 +176,16 @@ define i32 @callee_float_on_stack(i64 %a, i64 %b, i64 %c, i64 %d, float %e) {
 ;
 ; ILP32E-WITHFP-SAVE-RESTORE-LABEL: callee_float_on_stack:
 ; ILP32E-WITHFP-SAVE-RESTORE:       # %bb.0:
-; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    call t0, __riscv_save_1
-; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    .cfi_def_cfa_offset 8
+; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    call t0, __riscv_save_2
+; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    .cfi_def_cfa_offset 12
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    .cfi_offset ra, -4
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    .cfi_offset s0, -8
-; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    addi s0, sp, 8
+; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    addi s0, sp, 12
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    .cfi_def_cfa s0, 0
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    lw a0, 8(s0)
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    lw a1, 0(s0)
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    add a0, a1, a0
-; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    tail __riscv_restore_1
+; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    tail __riscv_restore_2
   %1 = trunc i64 %d to i32
   %2 = bitcast float %e to i32
   %3 = add i32 %1, %2
@@ -244,9 +244,9 @@ define i32 @caller_float_on_stack() {
 ;
 ; ILP32E-FPELIM-SAVE-RESTORE-LABEL: caller_float_on_stack:
 ; ILP32E-FPELIM-SAVE-RESTORE:       # %bb.0:
-; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    call t0, __riscv_save_0
+; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    call t0, __riscv_save_2
 ; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    addi sp, sp, -12
-; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    .cfi_def_cfa_offset 16
+; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    .cfi_def_cfa_offset 24
 ; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    .cfi_offset ra, -4
 ; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    lui a0, 264704
 ; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    sw a0, 8(sp)
@@ -261,16 +261,16 @@ define i32 @caller_float_on_stack() {
 ; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    li a5, 0
 ; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    call callee_float_on_stack
 ; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    addi sp, sp, 12
-; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    tail __riscv_restore_0
+; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    tail __riscv_restore_2
 ;
 ; ILP32E-WITHFP-SAVE-RESTORE-LABEL: caller_float_on_stack:
 ; ILP32E-WITHFP-SAVE-RESTORE:       # %bb.0:
-; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    call t0, __riscv_save_1
+; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    call t0, __riscv_save_2
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    addi sp, sp, -12
-; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    .cfi_def_cfa_offset 20
+; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    .cfi_def_cfa_offset 24
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    .cfi_offset ra, -4
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    .cfi_offset s0, -8
-; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    addi s0, sp, 20
+; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    addi s0, sp, 24
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    .cfi_def_cfa s0, 0
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    lui a0, 264704
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    sw a0, 8(sp)
@@ -285,7 +285,7 @@ define i32 @caller_float_on_stack() {
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    li a5, 0
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    call callee_float_on_stack
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    addi sp, sp, 12
-; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    tail __riscv_restore_1
+; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    tail __riscv_restore_2
   %1 = call i32 @callee_float_on_stack(i64 1, i64 2, i64 3, i64 4, float 5.0)
   ret i32 %1
 }
@@ -319,14 +319,14 @@ define float @callee_tiny_scalar_ret() {
 ;
 ; ILP32E-WITHFP-SAVE-RESTORE-LABEL: callee_tiny_scalar_ret:
 ; ILP32E-WITHFP-SAVE-RESTORE:       # %bb.0:
-; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    call t0, __riscv_save_1
-; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    .cfi_def_cfa_offset 8
+; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    call t0, __riscv_save_2
+; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    .cfi_def_cfa_offset 12
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    .cfi_offset ra, -4
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    .cfi_offset s0, -8
-; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    addi s0, sp, 8
+; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    addi s0, sp, 12
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    .cfi_def_cfa s0, 0
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    lui a0, 260096
-; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    tail __riscv_restore_1
+; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    tail __riscv_restore_2
   ret float 1.0
 }
 
@@ -360,22 +360,22 @@ define i32 @caller_tiny_scalar_ret() {
 ;
 ; ILP32E-FPELIM-SAVE-RESTORE-LABEL: caller_tiny_scalar_ret:
 ; ILP32E-FPELIM-SAVE-RESTORE:       # %bb.0:
-; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    call t0, __riscv_save_0
-; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    .cfi_def_cfa_offset 4
+; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    call t0, __riscv_save_2
+; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    .cfi_def_cfa_offset 12
 ; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    .cfi_offset ra, -4
 ; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    call callee_tiny_scalar_ret
-; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    tail __riscv_restore_0
+; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    tail __riscv_restore_2
 ;
 ; ILP32E-WITHFP-SAVE-RESTORE-LABEL: caller_tiny_scalar_ret:
 ; ILP32E-WITHFP-SAVE-RESTORE:       # %bb.0:
-; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    call t0, __riscv_save_1
-; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    .cfi_def_cfa_offset 8
+; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    call t0, __riscv_save_2
+; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    .cfi_def_cfa_offset 12
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    .cfi_offset ra, -4
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    .cfi_offset s0, -8
-; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    addi s0, sp, 8
+; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    addi s0, sp, 12
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    .cfi_def_cfa s0, 0
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    call callee_tiny_scalar_ret
-; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    tail __riscv_restore_1
+; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    tail __riscv_restore_2
   %1 = call float @callee_tiny_scalar_ret()
   %2 = bitcast float %1 to i32
   ret i32 %2
@@ -428,8 +428,8 @@ define i32 @callee_double_in_regs(i32 %a, double %b) {
 ;
 ; ILP32E-FPELIM-SAVE-RESTORE-LABEL: callee_double_in_regs:
 ; ILP32E-FPELIM-SAVE-RESTORE:       # %bb.0:
-; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    call t0, __riscv_save_1
-; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    .cfi_def_cfa_offset 8
+; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    call t0, __riscv_save_2
+; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    .cfi_def_cfa_offset 12
 ; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    .cfi_offset ra, -4
 ; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    .cfi_offset s0, -8
 ; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    mv s0, a0
@@ -437,7 +437,7 @@ define i32 @callee_double_in_regs(i32 %a, double %b) {
 ; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    mv a1, a2
 ; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    call __fixdfsi
 ; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    add a0, s0, a0
-; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    tail __riscv_restore_1
+; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    tail __riscv_restore_2
 ;
 ; ILP32E-WITHFP-SAVE-RESTORE-LABEL: callee_double_in_regs:
 ; ILP32E-WITHFP-SAVE-RESTORE:       # %bb.0:
@@ -495,28 +495,28 @@ define i32 @caller_double_in_regs() {
 ;
 ; ILP32E-FPELIM-SAVE-RESTORE-LABEL: caller_double_in_regs:
 ; ILP32E-FPELIM-SAVE-RESTORE:       # %bb.0:
-; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    call t0, __riscv_save_0
-; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    .cfi_def_cfa_offset 4
+; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    call t0, __riscv_save_2
+; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    .cfi_def_cfa_offset 12
 ; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    .cfi_offset ra, -4
 ; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    li a0, 1
 ; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    lui a2, 262144
 ; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    li a1, 0
 ; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    call callee_double_in_regs
-; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    tail __riscv_restore_0
+; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    tail __riscv_restore_2
 ;
 ; ILP32E-WITHFP-SAVE-RESTORE-LABEL: caller_double_in_regs:
 ; ILP32E-WITHFP-SAVE-RESTORE:       # %bb.0:
-; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    call t0, __riscv_save_1
-; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    .cfi_def_cfa_offset 8
+; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    call t0, __riscv_save_2
+; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    .cfi_def_cfa_offset 12
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    .cfi_offset ra, -4
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    .cfi_offset s0, -8
-; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    addi s0, sp, 8
+; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    addi s0, sp, 12
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    .cfi_def_cfa s0, 0
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    li a0, 1
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    lui a2, 262144
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    li a1, 0
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    call callee_double_in_regs
-; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    tail __riscv_restore_1
+; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    tail __riscv_restore_2
   %1 = call i32 @callee_double_in_regs(i32 1, double 2.0)
   ret i32 %1
 }
@@ -584,11 +584,11 @@ define i32 @callee_aligned_stack(i32 %a, i32 %b, fp128 %c, i32 %d, i32 %e, i64 %
 ;
 ; ILP32E-WITHFP-SAVE-RESTORE-LABEL: callee_aligned_stack:
 ; ILP32E-WITHFP-SAVE-RESTORE:       # %bb.0:
-; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    call t0, __riscv_save_1
-; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    .cfi_def_cfa_offset 8
+; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    call t0, __riscv_save_2
+; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    .cfi_def_cfa_offset 12
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    .cfi_offset ra, -4
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    .cfi_offset s0, -8
-; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    addi s0, sp, 8
+; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    addi s0, sp, 12
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    .cfi_def_cfa s0, 0
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    lw a0, 0(a2)
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    lw a1, 12(s0)
@@ -601,7 +601,7 @@ define i32 @callee_aligned_stack(i32 %a, i32 %b, fp128 %c, i32 %d, i32 %e, i64 %
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    add a0, a0, a1
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    add a4, a5, a4
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    add a0, a0, a4
-; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    tail __riscv_restore_1
+; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    tail __riscv_restore_2
   %1 = bitcast fp128 %c to i128
   %2 = trunc i128 %1 to i32
   %3 = add i32 %2, %g
@@ -730,8 +730,8 @@ define void @caller_aligned_stack() {
 ;
 ; ILP32E-FPELIM-SAVE-RESTORE-LABEL: caller_aligned_stack:
 ; ILP32E-FPELIM-SAVE-RESTORE:       # %bb.0:
-; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    call t0, __riscv_save_1
-; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    addi sp, sp, -56
+; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    call t0, __riscv_save_2
+; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    addi sp, sp, -52
 ; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    .cfi_def_cfa_offset 64
 ; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    .cfi_offset ra, -4
 ; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    .cfi_offset s0, -8
@@ -777,13 +777,13 @@ define void @caller_aligned_stack() {
 ; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    sw a6, 32(sp)
 ; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    call callee_aligned_stack
 ; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    addi sp, s0, -64
-; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    addi sp, sp, 56
-; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    tail __riscv_restore_1
+; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    addi sp, sp, 52
+; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    tail __riscv_restore_2
 ;
 ; ILP32E-WITHFP-SAVE-RESTORE-LABEL: caller_aligned_stack:
 ; ILP32E-WITHFP-SAVE-RESTORE:       # %bb.0:
-; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    call t0, __riscv_save_1
-; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    addi sp, sp, -56
+; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    call t0, __riscv_save_2
+; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    addi sp, sp, -52
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    .cfi_def_cfa_offset 64
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    .cfi_offset ra, -4
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    .cfi_offset s0, -8
@@ -829,8 +829,8 @@ define void @caller_aligned_stack() {
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    sw a6, 32(sp)
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    call callee_aligned_stack
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    addi sp, s0, -64
-; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    addi sp, sp, 56
-; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    tail __riscv_restore_1
+; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    addi sp, sp, 52
+; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    tail __riscv_restore_2
   %1 = call i32 @callee_aligned_stack(i32 1, i32 11,
     fp128 0xLEB851EB851EB851F400091EB851EB851, i32 12, i32 13,
     i64 20000000000, i32 14, i32 15, double 2.720000e+00, i32 16,
@@ -870,15 +870,15 @@ define double @callee_small_scalar_ret() {
 ;
 ; ILP32E-WITHFP-SAVE-RESTORE-LABEL: callee_small_scalar_ret:
 ; ILP32E-WITHFP-SAVE-RESTORE:       # %bb.0:
-; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    call t0, __riscv_save_1
-; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    .cfi_def_cfa_offset 8
+; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    call t0, __riscv_save_2
+; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    .cfi_def_cfa_offset 12
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    .cfi_offset ra, -4
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    .cfi_offset s0, -8
-; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    addi s0, sp, 8
+; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    addi s0, sp, 12
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    .cfi_def_cfa s0, 0
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    lui a1, 261888
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    li a0, 0
-; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    tail __riscv_restore_1
+; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    tail __riscv_restore_2
   ret double 1.0
 }
 
@@ -912,22 +912,22 @@ define i64 @caller_small_scalar_ret() {
 ;
 ; ILP32E-FPELIM-SAVE-RESTORE-LABEL: caller_small_scalar_ret:
 ; ILP32E-FPELIM-SAVE-RESTORE:       # %bb.0:
-; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    call t0, __riscv_save_0
-; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    .cfi_def_cfa_offset 4
+; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    call t0, __riscv_save_2
+; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    .cfi_def_cfa_offset 12
 ; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    .cfi_offset ra, -4
 ; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    call callee_small_scalar_ret
-; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    tail __riscv_restore_0
+; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    tail __riscv_restore_2
 ;
 ; ILP32E-WITHFP-SAVE-RESTORE-LABEL: caller_small_scalar_ret:
 ; ILP32E-WITHFP-SAVE-RESTORE:       # %bb.0:
-; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    call t0, __riscv_save_1
-; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    .cfi_def_cfa_offset 8
+; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    call t0, __riscv_save_2
+; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    .cfi_def_cfa_offset 12
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    .cfi_offset ra, -4
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    .cfi_offset s0, -8
-; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    addi s0, sp, 8
+; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    addi s0, sp, 12
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    .cfi_def_cfa s0, 0
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    call callee_small_scalar_ret
-; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    tail __riscv_restore_1
+; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    tail __riscv_restore_2
   %1 = call double @callee_small_scalar_ret()
   %2 = bitcast double %1 to i64
   ret i64 %2
@@ -965,14 +965,14 @@ define i32 @callee_i64_in_regs(i32 %a, i64 %b) {
 ;
 ; ILP32E-WITHFP-SAVE-RESTORE-LABEL: callee_i64_in_regs:
 ; ILP32E-WITHFP-SAVE-RESTORE:       # %bb.0:
-; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    call t0, __riscv_save_1
-; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    .cfi_def_cfa_offset 8
+; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    call t0, __riscv_save_2
+; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    .cfi_def_cfa_offset 12
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    .cfi_offset ra, -4
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    .cfi_offset s0, -8
-; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    addi s0, sp, 8
+; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    addi s0, sp, 12
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    .cfi_def_cfa s0, 0
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    add a0, a0, a1
-; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    tail __riscv_restore_1
+; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    tail __riscv_restore_2
   %b_trunc = trunc i64 %b to i32
   %1 = add i32 %a, %b_trunc
   ret i32 %1
@@ -1014,28 +1014,28 @@ define i32 @caller_i64_in_regs() {
 ;
 ; ILP32E-FPELIM-SAVE-RESTORE-LABEL: caller_i64_in_regs:
 ; ILP32E-FPELIM-SAVE-RESTORE:       # %bb.0:
-; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    call t0, __riscv_save_0
-; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    .cfi_def_cfa_offset 4
+; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    call t0, __riscv_save_2
+; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    .cfi_def_cfa_offset 12
 ; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    .cfi_offset ra, -4
 ; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    li a0, 1
 ; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    li a1, 2
 ; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    li a2, 0
 ; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    call callee_i64_in_regs
-; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    tail __riscv_restore_0
+; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    tail __riscv_restore_2
 ;
 ; ILP32E-WITHFP-SAVE-RESTORE-LABEL: caller_i64_in_regs:
 ; ILP32E-WITHFP-SAVE-RESTORE:       # %bb.0:
-; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    call t0, __riscv_save_1
-; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    .cfi_def_cfa_offset 8
+; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    call t0, __riscv_save_2
+; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    .cfi_def_cfa_offset 12
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    .cfi_offset ra, -4
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    .cfi_offset s0, -8
-; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    addi s0, sp, 8
+; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    addi s0, sp, 12
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    .cfi_def_cfa s0, 0
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    li a0, 1
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    li a1, 2
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    li a2, 0
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    call callee_i64_in_regs
-; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    tail __riscv_restore_1
+; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    tail __riscv_restore_2
   %1 = call i32 @callee_i64_in_regs(i32 1, i64 2)
   ret i32 %1
 }
@@ -1119,11 +1119,11 @@ define i32 @callee_many_scalars(i8 %a, i16 %b, i32 %c, i64 %d, i32 %e, i32 %f, i
 ;
 ; ILP32E-WITHFP-SAVE-RESTORE-LABEL: callee_many_scalars:
 ; ILP32E-WITHFP-SAVE-RESTORE:       # %bb.0:
-; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    call t0, __riscv_save_1
-; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    .cfi_def_cfa_offset 8
+; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    call t0, __riscv_save_2
+; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    .cfi_def_cfa_offset 12
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    .cfi_offset ra, -4
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    .cfi_offset s0, -8
-; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    addi s0, sp, 8
+; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    addi s0, sp, 12
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    .cfi_def_cfa s0, 0
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    lw a6, 12(s0)
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    lw a7, 0(s0)
@@ -1142,7 +1142,7 @@ define i32 @callee_many_scalars(i8 %a, i16 %b, i32 %c, i64 %d, i32 %e, i32 %f, i
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    add a0, a0, a7
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    add a0, a0, a6
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    add a0, a1, a0
-; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    tail __riscv_restore_1
+; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    tail __riscv_restore_2
   %a_ext = zext i8 %a to i32
   %b_ext = zext i16 %b to i32
   %1 = add i32 %a_ext, %b_ext
@@ -1212,9 +1212,9 @@ define i32 @caller_many_scalars() {
 ;
 ; ILP32E-FPELIM-SAVE-RESTORE-LABEL: caller_many_scalars:
 ; ILP32E-FPELIM-SAVE-RESTORE:       # %bb.0:
-; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    call t0, __riscv_save_0
+; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    call t0, __riscv_save_2
 ; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    addi sp, sp, -16
-; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    .cfi_def_cfa_offset 20
+; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    .cfi_def_cfa_offset 28
 ; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    .cfi_offset ra, -4
 ; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    li a0, 8
 ; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    sw a0, 12(sp)
@@ -1231,16 +1231,16 @@ define i32 @caller_many_scalars() {
 ; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    li a4, 0
 ; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    call callee_many_scalars
 ; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    addi sp, sp, 16
-; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    tail __riscv_restore_0
+; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    tail __riscv_restore_2
 ;
 ; ILP32E-WITHFP-SAVE-RESTORE-LABEL: caller_many_scalars:
 ; ILP32E-WITHFP-SAVE-RESTORE:       # %bb.0:
-; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    call t0, __riscv_save_1
+; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    call t0, __riscv_save_2
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    addi sp, sp, -16
-; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    .cfi_def_cfa_offset 24
+; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    .cfi_def_cfa_offset 28
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    .cfi_offset ra, -4
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    .cfi_offset s0, -8
-; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    addi s0, sp, 24
+; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    addi s0, sp, 28
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    .cfi_def_cfa s0, 0
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    li a0, 8
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    sw a0, 12(sp)
@@ -1257,7 +1257,7 @@ define i32 @caller_many_scalars() {
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    li a4, 0
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    call callee_many_scalars
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    addi sp, sp, 16
-; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    tail __riscv_restore_1
+; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    tail __riscv_restore_2
   %1 = call i32 @callee_many_scalars(i8 1, i16 2, i32 3, i64 4, i32 5, i32 6, i64 7, i32 8)
   ret i32 %1
 }
@@ -1338,11 +1338,11 @@ define i32 @callee_large_scalars(i128 %a, fp128 %b) {
 ;
 ; ILP32E-WITHFP-SAVE-RESTORE-LABEL: callee_large_scalars:
 ; ILP32E-WITHFP-SAVE-RESTORE:       # %bb.0:
-; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    call t0, __riscv_save_1
-; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    .cfi_def_cfa_offset 8
+; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    call t0, __riscv_save_2
+; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    .cfi_def_cfa_offset 12
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    .cfi_offset ra, -4
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    .cfi_offset s0, -8
-; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    addi s0, sp, 8
+; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    addi s0, sp, 12
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    .cfi_def_cfa s0, 0
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    lw a2, 0(a1)
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    lw a3, 0(a0)
@@ -1360,7 +1360,7 @@ define i32 @callee_large_scalars(i128 %a, fp128 %b) {
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    or a0, a2, a0
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    or a0, a0, a4
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    seqz a0, a0
-; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    tail __riscv_restore_1
+; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    tail __riscv_restore_2
   %b_bitcast = bitcast fp128 %b to i128
   %1 = icmp eq i128 %a, %b_bitcast
   %2 = zext i1 %1 to i32
@@ -1430,8 +1430,8 @@ define i32 @caller_large_scalars() {
 ;
 ; ILP32E-FPELIM-SAVE-RESTORE-LABEL: caller_large_scalars:
 ; ILP32E-FPELIM-SAVE-RESTORE:       # %bb.0:
-; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    call t0, __riscv_save_1
-; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    addi sp, sp, -40
+; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    call t0, __riscv_save_2
+; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    addi sp, sp, -36
 ; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    .cfi_def_cfa_offset 48
 ; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    .cfi_offset ra, -4
 ; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    .cfi_offset s0, -8
@@ -1443,22 +1443,22 @@ define i32 @caller_large_scalars() {
 ; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    sw zero, 8(sp)
 ; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    sw zero, 4(sp)
 ; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    sw zero, 0(sp)
-; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    sw zero, 36(sp)
-; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    sw zero, 32(sp)
 ; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    sw zero, 28(sp)
+; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    sw zero, 24(sp)
+; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    sw zero, 20(sp)
 ; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    li a2, 1
-; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    addi a0, sp, 24
+; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    addi a0, sp, 16
 ; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    mv a1, sp
-; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    sw a2, 24(sp)
+; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    sw a2, 16(sp)
 ; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    call callee_large_scalars
 ; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    addi sp, s0, -48
-; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    addi sp, sp, 40
-; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    tail __riscv_restore_1
+; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    addi sp, sp, 36
+; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    tail __riscv_restore_2
 ;
 ; ILP32E-WITHFP-SAVE-RESTORE-LABEL: caller_large_scalars:
 ; ILP32E-WITHFP-SAVE-RESTORE:       # %bb.0:
-; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    call t0, __riscv_save_1
-; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    addi sp, sp, -40
+; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    call t0, __riscv_save_2
+; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    addi sp, sp, -36
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    .cfi_def_cfa_offset 48
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    .cfi_offset ra, -4
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    .cfi_offset s0, -8
@@ -1470,17 +1470,17 @@ define i32 @caller_large_scalars() {
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    sw zero, 8(sp)
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    sw zero, 4(sp)
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    sw zero, 0(sp)
-; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    sw zero, 36(sp)
-; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    sw zero, 32(sp)
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    sw zero, 28(sp)
+; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    sw zero, 24(sp)
+; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    sw zero, 20(sp)
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    li a2, 1
-; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    addi a0, sp, 24
+; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    addi a0, sp, 16
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    mv a1, sp
-; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    sw a2, 24(sp)
+; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    sw a2, 16(sp)
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    call callee_large_scalars
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    addi sp, s0, -48
-; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    addi sp, sp, 40
-; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    tail __riscv_restore_1
+; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    addi sp, sp, 36
+; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    tail __riscv_restore_2
   %1 = call i32 @callee_large_scalars(i128 1, fp128 0xL00000000000000007FFF000000000000)
   ret i32 %1
 }
@@ -1569,11 +1569,11 @@ define i32 @callee_large_scalars_exhausted_regs(i32 %a, i32 %b, i32 %c, i32 %d,
 ;
 ; ILP32E-WITHFP-SAVE-RESTORE-LABEL: callee_large_scalars_exhausted_regs:
 ; ILP32E-WITHFP-SAVE-RESTORE:       # %bb.0:
-; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    call t0, __riscv_save_1
-; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    .cfi_def_cfa_offset 8
+; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    call t0, __riscv_save_2
+; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    .cfi_def_cfa_offset 12
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    .cfi_offset ra, -4
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    .cfi_offset s0, -8
-; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    addi s0, sp, 8
+; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    addi s0, sp, 12
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    .cfi_def_cfa s0, 0
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    lw a0, 12(s0)
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    lw a1, 4(s0)
@@ -1593,7 +1593,7 @@ define i32 @callee_large_scalars_exhausted_regs(i32 %a, i32 %b, i32 %c, i32 %d,
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    or a0, a2, a0
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    or a0, a0, a4
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    seqz a0, a0
-; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    tail __riscv_restore_1
+; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    tail __riscv_restore_2
   %j_bitcast = bitcast fp128 %j to i128
   %1 = icmp eq i128 %h, %j_bitcast
   %2 = zext i1 %1 to i32
@@ -1687,8 +1687,8 @@ define i32 @caller_large_scalars_exhausted_regs() {
 ;
 ; ILP32E-FPELIM-SAVE-RESTORE-LABEL: caller_large_scalars_exhausted_regs:
 ; ILP32E-FPELIM-SAVE-RESTORE:       # %bb.0:
-; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    call t0, __riscv_save_1
-; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    addi sp, sp, -56
+; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    call t0, __riscv_save_2
+; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    addi sp, sp, -52
 ; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    .cfi_def_cfa_offset 64
 ; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    .cfi_offset ra, -4
 ; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    .cfi_offset s0, -8
@@ -1699,7 +1699,7 @@ define i32 @caller_large_scalars_exhausted_regs() {
 ; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    sw a0, 12(sp)
 ; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    li a0, 9
 ; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    sw a0, 8(sp)
-; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    addi a0, sp, 40
+; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    addi a0, sp, 32
 ; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    sw a0, 4(sp)
 ; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    li a0, 7
 ; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    sw a0, 0(sp)
@@ -1708,26 +1708,26 @@ define i32 @caller_large_scalars_exhausted_regs() {
 ; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    sw zero, 24(sp)
 ; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    sw zero, 20(sp)
 ; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    sw zero, 16(sp)
-; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    sw zero, 52(sp)
-; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    sw zero, 48(sp)
+; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    sw zero, 44(sp)
+; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    sw zero, 40(sp)
 ; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    li a0, 8
-; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    sw a0, 40(sp)
+; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    sw a0, 32(sp)
 ; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    li a0, 1
 ; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    li a1, 2
 ; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    li a2, 3
 ; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    li a3, 4
 ; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    li a4, 5
 ; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    li a5, 6
-; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    sw zero, 44(sp)
+; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    sw zero, 36(sp)
 ; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    call callee_large_scalars_exhausted_regs
 ; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    addi sp, s0, -64
-; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    addi sp, sp, 56
-; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    tail __riscv_restore_1
+; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    addi sp, sp, 52
+; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    tail __riscv_restore_2
 ;
 ; ILP32E-WITHFP-SAVE-RESTORE-LABEL: caller_large_scalars_exhausted_regs:
 ; ILP32E-WITHFP-SAVE-RESTORE:       # %bb.0:
-; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    call t0, __riscv_save_1
-; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    addi sp, sp, -56
+; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    call t0, __riscv_save_2
+; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    addi sp, sp, -52
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    .cfi_def_cfa_offset 64
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    .cfi_offset ra, -4
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    .cfi_offset s0, -8
@@ -1738,7 +1738,7 @@ define i32 @caller_large_scalars_exhausted_regs() {
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    sw a0, 12(sp)
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    li a0, 9
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    sw a0, 8(sp)
-; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    addi a0, sp, 40
+; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    addi a0, sp, 32
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    sw a0, 4(sp)
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    li a0, 7
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    sw a0, 0(sp)
@@ -1747,21 +1747,21 @@ define i32 @caller_large_scalars_exhausted_regs() {
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    sw zero, 24(sp)
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    sw zero, 20(sp)
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    sw zero, 16(sp)
-; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    sw zero, 52(sp)
-; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    sw zero, 48(sp)
+; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    sw zero, 44(sp)
+; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    sw zero, 40(sp)
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    li a0, 8
-; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    sw a0, 40(sp)
+; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    sw a0, 32(sp)
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    li a0, 1
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    li a1, 2
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    li a2, 3
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    li a3, 4
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    li a4, 5
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    li a5, 6
-; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    sw zero, 44(sp)
+; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    sw zero, 36(sp)
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    call callee_large_scalars_exhausted_regs
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    addi sp, s0, -64
-; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    addi sp, sp, 56
-; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    tail __riscv_restore_1
+; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    addi sp, sp, 52
+; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    tail __riscv_restore_2
   %1 = call i32 @callee_large_scalars_exhausted_regs(
       i32 1, i32 2, i32 3, i32 4, i32 5, i32 6, i32 7, i128 8, i32 9,
       fp128 0xL00000000000000007FFF000000000000)
@@ -1817,12 +1817,12 @@ define i32 @caller_mixed_scalar_libcalls(i64 %a) {
 ;
 ; ILP32E-FPELIM-SAVE-RESTORE-LABEL: caller_mixed_scalar_libcalls:
 ; ILP32E-FPELIM-SAVE-RESTORE:       # %bb.0:
-; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    call t0, __riscv_save_1
-; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    addi sp, sp, -16
-; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    .cfi_def_cfa_offset 24
+; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    call t0, __riscv_save_2
+; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    addi sp, sp, -20
+; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    .cfi_def_cfa_offset 32
 ; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    .cfi_offset ra, -4
 ; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    .cfi_offset s0, -8
-; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    addi s0, sp, 24
+; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    addi s0, sp, 32
 ; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    .cfi_def_cfa s0, 0
 ; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    andi sp, sp, -8
 ; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    mv a2, a1
@@ -1830,18 +1830,18 @@ define i32 @caller_mixed_scalar_libcalls(i64 %a) {
 ; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    mv a0, sp
 ; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    call __floatditf
 ; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    lw a0, 0(sp)
-; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    addi sp, s0, -24
-; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    addi sp, sp, 16
-; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    tail __riscv_restore_1
+; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    addi sp, s0, -32
+; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    addi sp, sp, 20
+; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    tail __riscv_restore_2
 ;
 ; ILP32E-WITHFP-SAVE-RESTORE-LABEL: caller_mixed_scalar_libcalls:
 ; ILP32E-WITHFP-SAVE-RESTORE:       # %bb.0:
-; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    call t0, __riscv_save_1
-; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    addi sp, sp, -16
-; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    .cfi_def_cfa_offset 24
+; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    call t0, __riscv_save_2
+; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    addi sp, sp, -20
+; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    .cfi_def_cfa_offset 32
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    .cfi_offset ra, -4
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    .cfi_offset s0, -8
-; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    addi s0, sp, 24
+; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    addi s0, sp, 32
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    .cfi_def_cfa s0, 0
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    andi sp, sp, -8
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    mv a2, a1
@@ -1849,9 +1849,9 @@ define i32 @caller_mixed_scalar_libcalls(i64 %a) {
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    mv a0, sp
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    call __floatditf
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    lw a0, 0(sp)
-; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    addi sp, s0, -24
-; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    addi sp, sp, 16
-; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    tail __riscv_restore_1
+; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    addi sp, s0, -32
+; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    addi sp, sp, 20
+; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    tail __riscv_restore_2
   %1 = sitofp i64 %a to fp128
   %2 = bitcast fp128 %1 to i128
   %3 = trunc i128 %2 to i32
@@ -1895,15 +1895,15 @@ define i32 @callee_small_coerced_struct([2 x i32] %a.coerce) {
 ;
 ; ILP32E-WITHFP-SAVE-RESTORE-LABEL: callee_small_coerced_struct:
 ; ILP32E-WITHFP-SAVE-RESTORE:       # %bb.0:
-; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    call t0, __riscv_save_1
-; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    .cfi_def_cfa_offset 8
+; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    call t0, __riscv_save_2
+; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    .cfi_def_cfa_offset 12
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    .cfi_offset ra, -4
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    .cfi_offset s0, -8
-; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    addi s0, sp, 8
+; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    addi s0, sp, 12
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    .cfi_def_cfa s0, 0
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    xor a0, a0, a1
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    seqz a0, a0
-; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    tail __riscv_restore_1
+; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    tail __riscv_restore_2
   %1 = extractvalue [2 x i32] %a.coerce, 0
   %2 = extractvalue [2 x i32] %a.coerce, 1
   %3 = icmp eq i32 %1, %2
@@ -1945,26 +1945,26 @@ define i32 @caller_small_coerced_struct() {
 ;
 ; ILP32E-FPELIM-SAVE-RESTORE-LABEL: caller_small_coerced_struct:
 ; ILP32E-FPELIM-SAVE-RESTORE:       # %bb.0:
-; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    call t0, __riscv_save_0
-; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    .cfi_def_cfa_offset 4
+; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    call t0, __riscv_save_2
+; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    .cfi_def_cfa_offset 12
 ; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    .cfi_offset ra, -4
 ; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    li a0, 1
 ; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    li a1, 2
 ; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    call callee_small_coerced_struct
-; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    tail __riscv_restore_0
+; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    tail __riscv_restore_2
 ;
 ; ILP32E-WITHFP-SAVE-RESTORE-LABEL: caller_small_coerced_struct:
 ; ILP32E-WITHFP-SAVE-RESTORE:       # %bb.0:
-; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    call t0, __riscv_save_1
-; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    .cfi_def_cfa_offset 8
+; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    call t0, __riscv_save_2
+; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    .cfi_def_cfa_offset 12
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    .cfi_offset ra, -4
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    .cfi_offset s0, -8
-; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    addi s0, sp, 8
+; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    addi s0, sp, 12
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    .cfi_def_cfa s0, 0
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    li a0, 1
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    li a1, 2
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    call callee_small_coerced_struct
-; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    tail __riscv_restore_1
+; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    tail __riscv_restore_2
   %1 = call i32 @callee_small_coerced_struct([2 x i32] [i32 1, i32 2])
   ret i32 %1
 }
@@ -2008,16 +2008,16 @@ define i32 @callee_large_struct(ptr byval(%struct.large) align 4 %a) {
 ;
 ; ILP32E-WITHFP-SAVE-RESTORE-LABEL: callee_large_struct:
 ; ILP32E-WITHFP-SAVE-RESTORE:       # %bb.0:
-; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    call t0, __riscv_save_1
-; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    .cfi_def_cfa_offset 8
+; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    call t0, __riscv_save_2
+; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    .cfi_def_cfa_offset 12
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    .cfi_offset ra, -4
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    .cfi_offset s0, -8
-; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    addi s0, sp, 8
+; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    addi s0, sp, 12
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    .cfi_def_cfa s0, 0
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    lw a1, 0(a0)
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    lw a0, 12(a0)
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    add a0, a1, a0
-; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    tail __riscv_restore_1
+; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    tail __riscv_restore_2
   %1 = getelementptr inbounds %struct.large, ptr %a, i32 0, i32 0
   %2 = getelementptr inbounds %struct.large, ptr %a, i32 0, i32 3
   %3 = load i32, ptr %1
@@ -2082,9 +2082,9 @@ define i32 @caller_large_struct() {
 ;
 ; ILP32E-FPELIM-SAVE-RESTORE-LABEL: caller_large_struct:
 ; ILP32E-FPELIM-SAVE-RESTORE:       # %bb.0:
-; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    call t0, __riscv_save_0
+; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    call t0, __riscv_save_2
 ; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    addi sp, sp, -32
-; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    .cfi_def_cfa_offset 36
+; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    .cfi_def_cfa_offset 44
 ; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    .cfi_offset ra, -4
 ; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    li a0, 1
 ; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    sw a0, 16(sp)
@@ -2101,33 +2101,33 @@ define i32 @caller_large_struct() {
 ; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    mv a0, sp
 ; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    call callee_large_struct
 ; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    addi sp, sp, 32
-; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    tail __riscv_restore_0
+; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    tail __riscv_restore_2
 ;
 ; ILP32E-WITHFP-SAVE-RESTORE-LABEL: caller_large_struct:
 ; ILP32E-WITHFP-SAVE-RESTORE:       # %bb.0:
-; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    call t0, __riscv_save_1
+; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    call t0, __riscv_save_2
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    addi sp, sp, -32
-; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    .cfi_def_cfa_offset 40
+; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    .cfi_def_cfa_offset 44
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    .cfi_offset ra, -4
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    .cfi_offset s0, -8
-; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    addi s0, sp, 40
+; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    addi s0, sp, 44
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    .cfi_def_cfa s0, 0
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    li a0, 1
-; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    sw a0, -24(s0)
+; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    sw a0, -28(s0)
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    li a1, 2
-; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    sw a1, -20(s0)
+; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    sw a1, -24(s0)
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    li a2, 3
-; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    sw a2, -16(s0)
+; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    sw a2, -20(s0)
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    li a3, 4
-; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    sw a3, -12(s0)
-; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    sw a0, -40(s0)
-; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    sw a1, -36(s0)
-; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    sw a2, -32(s0)
-; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    sw a3, -28(s0)
-; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    addi a0, s0, -40
+; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    sw a3, -16(s0)
+; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    sw a0, -44(s0)
+; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    sw a1, -40(s0)
+; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    sw a2, -36(s0)
+; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    sw a3, -32(s0)
+; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    addi a0, s0, -44
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    call callee_large_struct
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    addi sp, sp, 32
-; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    tail __riscv_restore_1
+; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    tail __riscv_restore_2
   %ls = alloca %struct.large, align 4
   %1 = bitcast ptr %ls to ptr
   %a = getelementptr inbounds %struct.large, ptr %ls, i32 0, i32 0
@@ -2176,15 +2176,15 @@ define %struct.small @callee_small_struct_ret() {
 ;
 ; ILP32E-WITHFP-SAVE-RESTORE-LABEL: callee_small_struct_ret:
 ; ILP32E-WITHFP-SAVE-RESTORE:       # %bb.0:
-; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    call t0, __riscv_save_1
-; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    .cfi_def_cfa_offset 8
+; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    call t0, __riscv_save_2
+; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    .cfi_def_cfa_offset 12
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    .cfi_offset ra, -4
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    .cfi_offset s0, -8
-; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    addi s0, sp, 8
+; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    addi s0, sp, 12
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    .cfi_def_cfa s0, 0
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    li a0, 1
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    li a1, 0
-; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    tail __riscv_restore_1
+; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    tail __riscv_restore_2
   ret %struct.small { i32 1, ptr null }
 }
 
@@ -2220,24 +2220,24 @@ define i32 @caller_small_struct_ret() {
 ;
 ; ILP32E-FPELIM-SAVE-RESTORE-LABEL: caller_small_struct_ret:
 ; ILP32E-FPELIM-SAVE-RESTORE:       # %bb.0:
-; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    call t0, __riscv_save_0
-; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    .cfi_def_cfa_offset 4
+; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    call t0, __riscv_save_2
+; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    .cfi_def_cfa_offset 12
 ; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    .cfi_offset ra, -4
 ; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    call callee_small_struct_ret
 ; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    add a0, a0, a1
-; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    tail __riscv_restore_0
+; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    tail __riscv_restore_2
 ;
 ; ILP32E-WITHFP-SAVE-RESTORE-LABEL: caller_small_struct_ret:
 ; ILP32E-WITHFP-SAVE-RESTORE:       # %bb.0:
-; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    call t0, __riscv_save_1
-; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    .cfi_def_cfa_offset 8
+; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    call t0, __riscv_save_2
+; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    .cfi_def_cfa_offset 12
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    .cfi_offset ra, -4
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    .cfi_offset s0, -8
-; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    addi s0, sp, 8
+; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    addi s0, sp, 12
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    .cfi_def_cfa s0, 0
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    call callee_small_struct_ret
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    add a0, a0, a1
-; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    tail __riscv_restore_1
+; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    tail __riscv_restore_2
   %1 = call %struct.small @callee_small_struct_ret()
   %2 = extractvalue %struct.small %1, 0
   %3 = extractvalue %struct.small %1, 1
@@ -2289,18 +2289,18 @@ define fp128 @callee_large_scalar_ret() {
 ;
 ; ILP32E-WITHFP-SAVE-RESTORE-LABEL: callee_large_scalar_ret:
 ; ILP32E-WITHFP-SAVE-RESTORE:       # %bb.0:
-; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    call t0, __riscv_save_1
-; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    .cfi_def_cfa_offset 8
+; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    call t0, __riscv_save_2
+; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    .cfi_def_cfa_offset 12
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    .cfi_offset ra, -4
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    .cfi_offset s0, -8
-; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    addi s0, sp, 8
+; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    addi s0, sp, 12
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    .cfi_def_cfa s0, 0
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    lui a1, 524272
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    sw a1, 12(a0)
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    sw zero, 8(a0)
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    sw zero, 4(a0)
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    sw zero, 0(a0)
-; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    tail __riscv_restore_1
+; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    tail __riscv_restore_2
   ret fp128 0xL00000000000000007FFF000000000000
 }
 
@@ -2345,8 +2345,8 @@ define void @caller_large_scalar_ret() {
 ;
 ; ILP32E-FPELIM-SAVE-RESTORE-LABEL: caller_large_scalar_ret:
 ; ILP32E-FPELIM-SAVE-RESTORE:       # %bb.0:
-; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    call t0, __riscv_save_1
-; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    addi sp, sp, -24
+; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    call t0, __riscv_save_2
+; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    addi sp, sp, -20
 ; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    .cfi_def_cfa_offset 32
 ; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    .cfi_offset ra, -4
 ; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    .cfi_offset s0, -8
@@ -2356,13 +2356,13 @@ define void @caller_large_scalar_ret() {
 ; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    mv a0, sp
 ; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    call callee_large_scalar_ret
 ; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    addi sp, s0, -32
-; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    addi sp, sp, 24
-; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    tail __riscv_restore_1
+; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    addi sp, sp, 20
+; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    tail __riscv_restore_2
 ;
 ; ILP32E-WITHFP-SAVE-RESTORE-LABEL: caller_large_scalar_ret:
 ; ILP32E-WITHFP-SAVE-RESTORE:       # %bb.0:
-; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    call t0, __riscv_save_1
-; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    addi sp, sp, -24
+; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    call t0, __riscv_save_2
+; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    addi sp, sp, -20
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    .cfi_def_cfa_offset 32
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    .cfi_offset ra, -4
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    .cfi_offset s0, -8
@@ -2372,8 +2372,8 @@ define void @caller_large_scalar_ret() {
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    mv a0, sp
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    call callee_large_scalar_ret
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    addi sp, s0, -32
-; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    addi sp, sp, 24
-; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    tail __riscv_restore_1
+; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    addi sp, sp, 20
+; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    tail __riscv_restore_2
   %1 = call fp128 @callee_large_scalar_ret()
   ret void
 }
@@ -2430,11 +2430,11 @@ define void @callee_large_struct_ret(ptr noalias sret(%struct.large) %agg.result
 ;
 ; ILP32E-WITHFP-SAVE-RESTORE-LABEL: callee_large_struct_ret:
 ; ILP32E-WITHFP-SAVE-RESTORE:       # %bb.0:
-; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    call t0, __riscv_save_1
-; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    .cfi_def_cfa_offset 8
+; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    call t0, __riscv_save_2
+; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    .cfi_def_cfa_offset 12
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    .cfi_offset ra, -4
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    .cfi_offset s0, -8
-; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    addi s0, sp, 8
+; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    addi s0, sp, 12
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    .cfi_def_cfa s0, 0
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    li a1, 1
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    sw a1, 0(a0)
@@ -2444,7 +2444,7 @@ define void @callee_large_struct_ret(ptr noalias sret(%struct.large) %agg.result
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    sw a1, 8(a0)
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    li a1, 4
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    sw a1, 12(a0)
-; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    tail __riscv_restore_1
+; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    tail __riscv_restore_2
   %a = getelementptr inbounds %struct.large, ptr %agg.result, i32 0, i32 0
   store i32 1, ptr %a, align 4
   %b = getelementptr inbounds %struct.large, ptr %agg.result, i32 0, i32 1
@@ -2503,12 +2503,12 @@ define i32 @caller_large_struct_ret() {
 ;
 ; ILP32E-FPELIM-SAVE-RESTORE-LABEL: caller_large_struct_ret:
 ; ILP32E-FPELIM-SAVE-RESTORE:       # %bb.0:
-; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    call t0, __riscv_save_1
-; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    addi sp, sp, -16
-; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    .cfi_def_cfa_offset 24
+; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    call t0, __riscv_save_2
+; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    addi sp, sp, -20
+; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    .cfi_def_cfa_offset 32
 ; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    .cfi_offset ra, -4
 ; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    .cfi_offset s0, -8
-; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    addi s0, sp, 24
+; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    addi s0, sp, 32
 ; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    .cfi_def_cfa s0, 0
 ; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    andi sp, sp, -8
 ; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    mv a0, sp
@@ -2516,18 +2516,18 @@ define i32 @caller_large_struct_ret() {
 ; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    lw a0, 0(sp)
 ; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    lw a1, 12(sp)
 ; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    add a0, a0, a1
-; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    addi sp, s0, -24
-; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    addi sp, sp, 16
-; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    tail __riscv_restore_1
+; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    addi sp, s0, -32
+; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    addi sp, sp, 20
+; ILP32E-FPELIM-SAVE-RESTORE-NEXT:    tail __riscv_restore_2
 ;
 ; ILP32E-WITHFP-SAVE-RESTORE-LABEL: caller_large_struct_ret:
 ; ILP32E-WITHFP-SAVE-RESTORE:       # %bb.0:
-; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    call t0, __riscv_save_1
-; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    addi sp, sp, -16
-; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    .cfi_def_cfa_offset 24
+; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    call t0, __riscv_save_2
+; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    addi sp, sp, -20
+; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    .cfi_def_cfa_offset 32
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    .cfi_offset ra, -4
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    .cfi_offset s0, -8
-; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    addi s0, sp, 24
+; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    addi s0, sp, 32
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    .cfi_def_cfa s0, 0
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    andi sp, sp, -8
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    mv a0, sp
@@ -2535,9 +2535,9 @@ define i32 @caller_large_struct_ret() {
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    lw a0, 0(sp)
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    lw a1, 12(sp)
 ; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    add a0, a0, a1
-; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    addi sp, s0, -24
-; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    addi sp, sp, 16
-; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    tail __riscv_restore_1
+; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    addi sp, s0, -32
+; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    addi sp, sp, 20
+; ILP32E-WITHFP-SAVE-RESTORE-NEXT:    tail __riscv_restore_2
   %1 = alloca %struct.large
   call void @callee_large_struct_ret(ptr sret(%struct.large) %1)
   %2 = getelementptr inbounds %struct.large, ptr %1, i32 0, i32 0