llvm/llvm/test/CodeGen/LoongArch/machinelicm-address-pseudos.ll

; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
; RUN: llc --mtriple=loongarch32 --relocation-model=pic \
; RUN:   --verify-machineinstrs < %s | FileCheck -check-prefixes=LA32 %s
; RUN: llc --mtriple=loongarch64 --relocation-model=pic \
; RUN:   --verify-machineinstrs < %s | FileCheck -check-prefixes=LA64 %s
; RUN: llc --mtriple=loongarch64 --relocation-model=pic --code-model=large \
; RUN:   --verify-machineinstrs < %s | FileCheck -check-prefixes=LA64LARGE %s

; Verifies that MachineLICM can hoist address generation pseudos out of loops.

@l = protected global i32 0, align 4

define void @test_la_pcrel(i32 signext %n) {
; LA32-LABEL: test_la_pcrel:
; LA32:       # %bb.0: # %entry
; LA32-NEXT:    move $a1, $zero
; LA32-NEXT:    pcalau12i $a2, %pc_hi20(l)
; LA32-NEXT:    .p2align 4, , 16
; LA32-NEXT:  .LBB0_1: # %loop
; LA32-NEXT:    # =>This Inner Loop Header: Depth=1
; LA32-NEXT:    ld.w $zero, $a2, %pc_lo12(l)
; LA32-NEXT:    addi.w $a1, $a1, 1
; LA32-NEXT:    blt $a1, $a0, .LBB0_1
; LA32-NEXT:  # %bb.2: # %ret
; LA32-NEXT:    ret
;
; LA64-LABEL: test_la_pcrel:
; LA64:       # %bb.0: # %entry
; LA64-NEXT:    move $a1, $zero
; LA64-NEXT:    pcalau12i $a2, %pc_hi20(l)
; LA64-NEXT:    .p2align 4, , 16
; LA64-NEXT:  .LBB0_1: # %loop
; LA64-NEXT:    # =>This Inner Loop Header: Depth=1
; LA64-NEXT:    ld.w $zero, $a2, %pc_lo12(l)
; LA64-NEXT:    addi.w $a1, $a1, 1
; LA64-NEXT:    blt $a1, $a0, .LBB0_1
; LA64-NEXT:  # %bb.2: # %ret
; LA64-NEXT:    ret
;
; LA64LARGE-LABEL: test_la_pcrel:
; LA64LARGE:       # %bb.0: # %entry
; LA64LARGE-NEXT:    pcalau12i $a1, %pc_hi20(l)
; LA64LARGE-NEXT:    addi.d $a2, $zero, %pc_lo12(l)
; LA64LARGE-NEXT:    lu32i.d $a2, %pc64_lo20(l)
; LA64LARGE-NEXT:    lu52i.d $a2, $a2, %pc64_hi12(l)
; LA64LARGE-NEXT:    move $a3, $zero
; LA64LARGE-NEXT:    .p2align 4, , 16
; LA64LARGE-NEXT:  .LBB0_1: # %loop
; LA64LARGE-NEXT:    # =>This Inner Loop Header: Depth=1
; LA64LARGE-NEXT:    ldx.w $zero, $a2, $a1
; LA64LARGE-NEXT:    addi.w $a3, $a3, 1
; LA64LARGE-NEXT:    blt $a3, $a0, .LBB0_1
; LA64LARGE-NEXT:  # %bb.2: # %ret
; LA64LARGE-NEXT:    ret
entry:
  br label %loop

loop:
  %i = phi i32 [ %inc, %loop ], [ 0, %entry ]
  %0 = load volatile i32, ptr @l, align 4
  %inc = add nuw nsw i32 %i, 1
  %cmp = icmp slt i32 %inc, %n
  br i1 %cmp, label %loop, label %ret

ret:
  ret void
}

@g = global i32 0, align 4

define void @test_la_got(i32 signext %n) {
; LA32-LABEL: test_la_got:
; LA32:       # %bb.0: # %entry
; LA32-NEXT:    pcalau12i $a1, %got_pc_hi20(g)
; LA32-NEXT:    ld.w $a1, $a1, %got_pc_lo12(g)
; LA32-NEXT:    move $a2, $zero
; LA32-NEXT:    .p2align 4, , 16
; LA32-NEXT:  .LBB1_1: # %loop
; LA32-NEXT:    # =>This Inner Loop Header: Depth=1
; LA32-NEXT:    ld.w $zero, $a1, 0
; LA32-NEXT:    addi.w $a2, $a2, 1
; LA32-NEXT:    blt $a2, $a0, .LBB1_1
; LA32-NEXT:  # %bb.2: # %ret
; LA32-NEXT:    ret
;
; LA64-LABEL: test_la_got:
; LA64:       # %bb.0: # %entry
; LA64-NEXT:    pcalau12i $a1, %got_pc_hi20(g)
; LA64-NEXT:    ld.d $a1, $a1, %got_pc_lo12(g)
; LA64-NEXT:    move $a2, $zero
; LA64-NEXT:    .p2align 4, , 16
; LA64-NEXT:  .LBB1_1: # %loop
; LA64-NEXT:    # =>This Inner Loop Header: Depth=1
; LA64-NEXT:    ld.w $zero, $a1, 0
; LA64-NEXT:    addi.w $a2, $a2, 1
; LA64-NEXT:    blt $a2, $a0, .LBB1_1
; LA64-NEXT:  # %bb.2: # %ret
; LA64-NEXT:    ret
;
; LA64LARGE-LABEL: test_la_got:
; LA64LARGE:       # %bb.0: # %entry
; LA64LARGE-NEXT:    pcalau12i $a1, %got_pc_hi20(g)
; LA64LARGE-NEXT:    addi.d $a2, $zero, %got_pc_lo12(g)
; LA64LARGE-NEXT:    lu32i.d $a2, %got64_pc_lo20(g)
; LA64LARGE-NEXT:    lu52i.d $a2, $a2, %got64_pc_hi12(g)
; LA64LARGE-NEXT:    ldx.d $a1, $a2, $a1
; LA64LARGE-NEXT:    move $a2, $zero
; LA64LARGE-NEXT:    .p2align 4, , 16
; LA64LARGE-NEXT:  .LBB1_1: # %loop
; LA64LARGE-NEXT:    # =>This Inner Loop Header: Depth=1
; LA64LARGE-NEXT:    ld.w $zero, $a1, 0
; LA64LARGE-NEXT:    addi.w $a2, $a2, 1
; LA64LARGE-NEXT:    blt $a2, $a0, .LBB1_1
; LA64LARGE-NEXT:  # %bb.2: # %ret
; LA64LARGE-NEXT:    ret
entry:
  br label %loop

loop:
  %i = phi i32 [ %inc, %loop ], [ 0, %entry ]
  %0 = load volatile i32, ptr @g, align 4
  %inc = add nuw nsw i32 %i, 1
  %cmp = icmp slt i32 %inc, %n
  br i1 %cmp, label %loop, label %ret

ret:
  ret void
}

@ie = external thread_local(initialexec) global i32

define void @test_la_tls_ie(i32 signext %n) {
; LA32-LABEL: test_la_tls_ie:
; LA32:       # %bb.0: # %entry
; LA32-NEXT:    pcalau12i $a1, %ie_pc_hi20(ie)
; LA32-NEXT:    ld.w $a2, $a1, %ie_pc_lo12(ie)
; LA32-NEXT:    move $a1, $zero
; LA32-NEXT:    add.w $a2, $a2, $tp
; LA32-NEXT:    .p2align 4, , 16
; LA32-NEXT:  .LBB2_1: # %loop
; LA32-NEXT:    # =>This Inner Loop Header: Depth=1
; LA32-NEXT:    ld.w $zero, $a2, 0
; LA32-NEXT:    addi.w $a1, $a1, 1
; LA32-NEXT:    blt $a1, $a0, .LBB2_1
; LA32-NEXT:  # %bb.2: # %ret
; LA32-NEXT:    ret
;
; LA64-LABEL: test_la_tls_ie:
; LA64:       # %bb.0: # %entry
; LA64-NEXT:    pcalau12i $a1, %ie_pc_hi20(ie)
; LA64-NEXT:    ld.d $a1, $a1, %ie_pc_lo12(ie)
; LA64-NEXT:    move $a2, $zero
; LA64-NEXT:    .p2align 4, , 16
; LA64-NEXT:  .LBB2_1: # %loop
; LA64-NEXT:    # =>This Inner Loop Header: Depth=1
; LA64-NEXT:    ldx.w $zero, $a1, $tp
; LA64-NEXT:    addi.w $a2, $a2, 1
; LA64-NEXT:    blt $a2, $a0, .LBB2_1
; LA64-NEXT:  # %bb.2: # %ret
; LA64-NEXT:    ret
;
; LA64LARGE-LABEL: test_la_tls_ie:
; LA64LARGE:       # %bb.0: # %entry
; LA64LARGE-NEXT:    pcalau12i $a1, %ie_pc_hi20(ie)
; LA64LARGE-NEXT:    addi.d $a2, $zero, %ie_pc_lo12(ie)
; LA64LARGE-NEXT:    lu32i.d $a2, %ie64_pc_lo20(ie)
; LA64LARGE-NEXT:    lu52i.d $a2, $a2, %ie64_pc_hi12(ie)
; LA64LARGE-NEXT:    ldx.d $a1, $a2, $a1
; LA64LARGE-NEXT:    move $a2, $zero
; LA64LARGE-NEXT:    .p2align 4, , 16
; LA64LARGE-NEXT:  .LBB2_1: # %loop
; LA64LARGE-NEXT:    # =>This Inner Loop Header: Depth=1
; LA64LARGE-NEXT:    ldx.w $zero, $a1, $tp
; LA64LARGE-NEXT:    addi.w $a2, $a2, 1
; LA64LARGE-NEXT:    blt $a2, $a0, .LBB2_1
; LA64LARGE-NEXT:  # %bb.2: # %ret
; LA64LARGE-NEXT:    ret
entry:
  br label %loop

loop:
  %i = phi i32 [ %inc, %loop ], [ 0, %entry ]
  %0 = load volatile i32, ptr @ie, align 4
  %inc = add nuw nsw i32 %i, 1
  %cmp = icmp slt i32 %inc, %n
  br i1 %cmp, label %loop, label %ret

ret:
  ret void
}

@ld = external thread_local(localdynamic) global i32

define void @test_la_tls_ld(i32 signext %n) {
; LA32-LABEL: test_la_tls_ld:
; LA32:       # %bb.0: # %entry
; LA32-NEXT:    addi.w $sp, $sp, -16
; LA32-NEXT:    .cfi_def_cfa_offset 16
; LA32-NEXT:    st.w $ra, $sp, 12 # 4-byte Folded Spill
; LA32-NEXT:    st.w $fp, $sp, 8 # 4-byte Folded Spill
; LA32-NEXT:    st.w $s0, $sp, 4 # 4-byte Folded Spill
; LA32-NEXT:    st.w $s1, $sp, 0 # 4-byte Folded Spill
; LA32-NEXT:    .cfi_offset 1, -4
; LA32-NEXT:    .cfi_offset 22, -8
; LA32-NEXT:    .cfi_offset 23, -12
; LA32-NEXT:    .cfi_offset 24, -16
; LA32-NEXT:    move $fp, $a0
; LA32-NEXT:    move $s1, $zero
; LA32-NEXT:    pcalau12i $a0, %ld_pc_hi20(ld)
; LA32-NEXT:    addi.w $s0, $a0, %got_pc_lo12(ld)
; LA32-NEXT:    .p2align 4, , 16
; LA32-NEXT:  .LBB3_1: # %loop
; LA32-NEXT:    # =>This Inner Loop Header: Depth=1
; LA32-NEXT:    move $a0, $s0
; LA32-NEXT:    bl %plt(__tls_get_addr)
; LA32-NEXT:    ld.w $zero, $a0, 0
; LA32-NEXT:    addi.w $s1, $s1, 1
; LA32-NEXT:    blt $s1, $fp, .LBB3_1
; LA32-NEXT:  # %bb.2: # %ret
; LA32-NEXT:    ld.w $s1, $sp, 0 # 4-byte Folded Reload
; LA32-NEXT:    ld.w $s0, $sp, 4 # 4-byte Folded Reload
; LA32-NEXT:    ld.w $fp, $sp, 8 # 4-byte Folded Reload
; LA32-NEXT:    ld.w $ra, $sp, 12 # 4-byte Folded Reload
; LA32-NEXT:    addi.w $sp, $sp, 16
; LA32-NEXT:    ret
;
; LA64-LABEL: test_la_tls_ld:
; LA64:       # %bb.0: # %entry
; LA64-NEXT:    addi.d $sp, $sp, -32
; LA64-NEXT:    .cfi_def_cfa_offset 32
; LA64-NEXT:    st.d $ra, $sp, 24 # 8-byte Folded Spill
; LA64-NEXT:    st.d $fp, $sp, 16 # 8-byte Folded Spill
; LA64-NEXT:    st.d $s0, $sp, 8 # 8-byte Folded Spill
; LA64-NEXT:    st.d $s1, $sp, 0 # 8-byte Folded Spill
; LA64-NEXT:    .cfi_offset 1, -8
; LA64-NEXT:    .cfi_offset 22, -16
; LA64-NEXT:    .cfi_offset 23, -24
; LA64-NEXT:    .cfi_offset 24, -32
; LA64-NEXT:    move $fp, $a0
; LA64-NEXT:    move $s1, $zero
; LA64-NEXT:    pcalau12i $a0, %ld_pc_hi20(ld)
; LA64-NEXT:    addi.d $s0, $a0, %got_pc_lo12(ld)
; LA64-NEXT:    .p2align 4, , 16
; LA64-NEXT:  .LBB3_1: # %loop
; LA64-NEXT:    # =>This Inner Loop Header: Depth=1
; LA64-NEXT:    move $a0, $s0
; LA64-NEXT:    bl %plt(__tls_get_addr)
; LA64-NEXT:    ld.w $zero, $a0, 0
; LA64-NEXT:    addi.w $s1, $s1, 1
; LA64-NEXT:    blt $s1, $fp, .LBB3_1
; LA64-NEXT:  # %bb.2: # %ret
; LA64-NEXT:    ld.d $s1, $sp, 0 # 8-byte Folded Reload
; LA64-NEXT:    ld.d $s0, $sp, 8 # 8-byte Folded Reload
; LA64-NEXT:    ld.d $fp, $sp, 16 # 8-byte Folded Reload
; LA64-NEXT:    ld.d $ra, $sp, 24 # 8-byte Folded Reload
; LA64-NEXT:    addi.d $sp, $sp, 32
; LA64-NEXT:    ret
;
; LA64LARGE-LABEL: test_la_tls_ld:
; LA64LARGE:       # %bb.0: # %entry
; LA64LARGE-NEXT:    addi.d $sp, $sp, -32
; LA64LARGE-NEXT:    .cfi_def_cfa_offset 32
; LA64LARGE-NEXT:    st.d $ra, $sp, 24 # 8-byte Folded Spill
; LA64LARGE-NEXT:    st.d $fp, $sp, 16 # 8-byte Folded Spill
; LA64LARGE-NEXT:    st.d $s0, $sp, 8 # 8-byte Folded Spill
; LA64LARGE-NEXT:    st.d $s1, $sp, 0 # 8-byte Folded Spill
; LA64LARGE-NEXT:    .cfi_offset 1, -8
; LA64LARGE-NEXT:    .cfi_offset 22, -16
; LA64LARGE-NEXT:    .cfi_offset 23, -24
; LA64LARGE-NEXT:    .cfi_offset 24, -32
; LA64LARGE-NEXT:    move $fp, $a0
; LA64LARGE-NEXT:    pcalau12i $a0, %ld_pc_hi20(ld)
; LA64LARGE-NEXT:    addi.d $a1, $zero, %got_pc_lo12(ld)
; LA64LARGE-NEXT:    lu32i.d $a1, %got64_pc_lo20(ld)
; LA64LARGE-NEXT:    lu52i.d $a1, $a1, %got64_pc_hi12(ld)
; LA64LARGE-NEXT:    move $s1, $zero
; LA64LARGE-NEXT:    add.d $s0, $a1, $a0
; LA64LARGE-NEXT:    .p2align 4, , 16
; LA64LARGE-NEXT:  .LBB3_1: # %loop
; LA64LARGE-NEXT:    # =>This Inner Loop Header: Depth=1
; LA64LARGE-NEXT:    move $a0, $s0
; LA64LARGE-NEXT:    pcalau12i $a1, %pc_hi20(__tls_get_addr)
; LA64LARGE-NEXT:    addi.d $ra, $zero, %pc_lo12(__tls_get_addr)
; LA64LARGE-NEXT:    lu32i.d $ra, %pc64_lo20(__tls_get_addr)
; LA64LARGE-NEXT:    lu52i.d $ra, $ra, %pc64_hi12(__tls_get_addr)
; LA64LARGE-NEXT:    add.d $ra, $ra, $a1
; LA64LARGE-NEXT:    jirl $ra, $ra, 0
; LA64LARGE-NEXT:    ld.w $zero, $a0, 0
; LA64LARGE-NEXT:    addi.w $s1, $s1, 1
; LA64LARGE-NEXT:    blt $s1, $fp, .LBB3_1
; LA64LARGE-NEXT:  # %bb.2: # %ret
; LA64LARGE-NEXT:    ld.d $s1, $sp, 0 # 8-byte Folded Reload
; LA64LARGE-NEXT:    ld.d $s0, $sp, 8 # 8-byte Folded Reload
; LA64LARGE-NEXT:    ld.d $fp, $sp, 16 # 8-byte Folded Reload
; LA64LARGE-NEXT:    ld.d $ra, $sp, 24 # 8-byte Folded Reload
; LA64LARGE-NEXT:    addi.d $sp, $sp, 32
; LA64LARGE-NEXT:    ret
entry:
  br label %loop

loop:
  %i = phi i32 [ %inc, %loop ], [ 0, %entry ]
  %0 = load volatile i32, ptr @ld, align 4
  %inc = add nuw nsw i32 %i, 1
  %cmp = icmp slt i32 %inc, %n
  br i1 %cmp, label %loop, label %ret

ret:
  ret void
}

@le = external thread_local(localexec) global i32

define void @test_la_tls_le(i32 signext %n) {
; LA32-LABEL: test_la_tls_le:
; LA32:       # %bb.0: # %entry
; LA32-NEXT:    move $a1, $zero
; LA32-NEXT:    lu12i.w $a2, %le_hi20(le)
; LA32-NEXT:    ori $a2, $a2, %le_lo12(le)
; LA32-NEXT:    add.w $a2, $a2, $tp
; LA32-NEXT:    .p2align 4, , 16
; LA32-NEXT:  .LBB4_1: # %loop
; LA32-NEXT:    # =>This Inner Loop Header: Depth=1
; LA32-NEXT:    ld.w $zero, $a2, 0
; LA32-NEXT:    addi.w $a1, $a1, 1
; LA32-NEXT:    blt $a1, $a0, .LBB4_1
; LA32-NEXT:  # %bb.2: # %ret
; LA32-NEXT:    ret
;
; LA64-LABEL: test_la_tls_le:
; LA64:       # %bb.0: # %entry
; LA64-NEXT:    move $a1, $zero
; LA64-NEXT:    lu12i.w $a2, %le_hi20(le)
; LA64-NEXT:    ori $a2, $a2, %le_lo12(le)
; LA64-NEXT:    .p2align 4, , 16
; LA64-NEXT:  .LBB4_1: # %loop
; LA64-NEXT:    # =>This Inner Loop Header: Depth=1
; LA64-NEXT:    ldx.w $zero, $a2, $tp
; LA64-NEXT:    addi.w $a1, $a1, 1
; LA64-NEXT:    blt $a1, $a0, .LBB4_1
; LA64-NEXT:  # %bb.2: # %ret
; LA64-NEXT:    ret
;
; LA64LARGE-LABEL: test_la_tls_le:
; LA64LARGE:       # %bb.0: # %entry
; LA64LARGE-NEXT:    move $a1, $zero
; LA64LARGE-NEXT:    lu12i.w $a2, %le_hi20(le)
; LA64LARGE-NEXT:    ori $a2, $a2, %le_lo12(le)
; LA64LARGE-NEXT:    lu32i.d $a2, %le64_lo20(le)
; LA64LARGE-NEXT:    lu52i.d $a2, $a2, %le64_hi12(le)
; LA64LARGE-NEXT:    .p2align 4, , 16
; LA64LARGE-NEXT:  .LBB4_1: # %loop
; LA64LARGE-NEXT:    # =>This Inner Loop Header: Depth=1
; LA64LARGE-NEXT:    ldx.w $zero, $a2, $tp
; LA64LARGE-NEXT:    addi.w $a1, $a1, 1
; LA64LARGE-NEXT:    blt $a1, $a0, .LBB4_1
; LA64LARGE-NEXT:  # %bb.2: # %ret
; LA64LARGE-NEXT:    ret
entry:
  br label %loop

loop:
  %i = phi i32 [ %inc, %loop ], [ 0, %entry ]
  %0 = load volatile i32, ptr @le, align 4
  %inc = add nuw nsw i32 %i, 1
  %cmp = icmp slt i32 %inc, %n
  br i1 %cmp, label %loop, label %ret

ret:
  ret void
}

@gd = external thread_local global i32

define void @test_la_tls_gd(i32 signext %n) nounwind {
; LA32-LABEL: test_la_tls_gd:
; LA32:       # %bb.0: # %entry
; LA32-NEXT:    addi.w $sp, $sp, -16
; LA32-NEXT:    st.w $ra, $sp, 12 # 4-byte Folded Spill
; LA32-NEXT:    st.w $fp, $sp, 8 # 4-byte Folded Spill
; LA32-NEXT:    st.w $s0, $sp, 4 # 4-byte Folded Spill
; LA32-NEXT:    st.w $s1, $sp, 0 # 4-byte Folded Spill
; LA32-NEXT:    move $fp, $a0
; LA32-NEXT:    move $s1, $zero
; LA32-NEXT:    pcalau12i $a0, %gd_pc_hi20(gd)
; LA32-NEXT:    addi.w $s0, $a0, %got_pc_lo12(gd)
; LA32-NEXT:    .p2align 4, , 16
; LA32-NEXT:  .LBB5_1: # %loop
; LA32-NEXT:    # =>This Inner Loop Header: Depth=1
; LA32-NEXT:    move $a0, $s0
; LA32-NEXT:    bl %plt(__tls_get_addr)
; LA32-NEXT:    ld.w $zero, $a0, 0
; LA32-NEXT:    addi.w $s1, $s1, 1
; LA32-NEXT:    blt $s1, $fp, .LBB5_1
; LA32-NEXT:  # %bb.2: # %ret
; LA32-NEXT:    ld.w $s1, $sp, 0 # 4-byte Folded Reload
; LA32-NEXT:    ld.w $s0, $sp, 4 # 4-byte Folded Reload
; LA32-NEXT:    ld.w $fp, $sp, 8 # 4-byte Folded Reload
; LA32-NEXT:    ld.w $ra, $sp, 12 # 4-byte Folded Reload
; LA32-NEXT:    addi.w $sp, $sp, 16
; LA32-NEXT:    ret
;
; LA64-LABEL: test_la_tls_gd:
; LA64:       # %bb.0: # %entry
; LA64-NEXT:    addi.d $sp, $sp, -32
; LA64-NEXT:    st.d $ra, $sp, 24 # 8-byte Folded Spill
; LA64-NEXT:    st.d $fp, $sp, 16 # 8-byte Folded Spill
; LA64-NEXT:    st.d $s0, $sp, 8 # 8-byte Folded Spill
; LA64-NEXT:    st.d $s1, $sp, 0 # 8-byte Folded Spill
; LA64-NEXT:    move $fp, $a0
; LA64-NEXT:    move $s1, $zero
; LA64-NEXT:    pcalau12i $a0, %gd_pc_hi20(gd)
; LA64-NEXT:    addi.d $s0, $a0, %got_pc_lo12(gd)
; LA64-NEXT:    .p2align 4, , 16
; LA64-NEXT:  .LBB5_1: # %loop
; LA64-NEXT:    # =>This Inner Loop Header: Depth=1
; LA64-NEXT:    move $a0, $s0
; LA64-NEXT:    bl %plt(__tls_get_addr)
; LA64-NEXT:    ld.w $zero, $a0, 0
; LA64-NEXT:    addi.w $s1, $s1, 1
; LA64-NEXT:    blt $s1, $fp, .LBB5_1
; LA64-NEXT:  # %bb.2: # %ret
; LA64-NEXT:    ld.d $s1, $sp, 0 # 8-byte Folded Reload
; LA64-NEXT:    ld.d $s0, $sp, 8 # 8-byte Folded Reload
; LA64-NEXT:    ld.d $fp, $sp, 16 # 8-byte Folded Reload
; LA64-NEXT:    ld.d $ra, $sp, 24 # 8-byte Folded Reload
; LA64-NEXT:    addi.d $sp, $sp, 32
; LA64-NEXT:    ret
;
; LA64LARGE-LABEL: test_la_tls_gd:
; LA64LARGE:       # %bb.0: # %entry
; LA64LARGE-NEXT:    addi.d $sp, $sp, -32
; LA64LARGE-NEXT:    st.d $ra, $sp, 24 # 8-byte Folded Spill
; LA64LARGE-NEXT:    st.d $fp, $sp, 16 # 8-byte Folded Spill
; LA64LARGE-NEXT:    st.d $s0, $sp, 8 # 8-byte Folded Spill
; LA64LARGE-NEXT:    st.d $s1, $sp, 0 # 8-byte Folded Spill
; LA64LARGE-NEXT:    move $fp, $a0
; LA64LARGE-NEXT:    pcalau12i $a0, %gd_pc_hi20(gd)
; LA64LARGE-NEXT:    addi.d $a1, $zero, %got_pc_lo12(gd)
; LA64LARGE-NEXT:    lu32i.d $a1, %got64_pc_lo20(gd)
; LA64LARGE-NEXT:    lu52i.d $a1, $a1, %got64_pc_hi12(gd)
; LA64LARGE-NEXT:    move $s1, $zero
; LA64LARGE-NEXT:    add.d $s0, $a1, $a0
; LA64LARGE-NEXT:    .p2align 4, , 16
; LA64LARGE-NEXT:  .LBB5_1: # %loop
; LA64LARGE-NEXT:    # =>This Inner Loop Header: Depth=1
; LA64LARGE-NEXT:    move $a0, $s0
; LA64LARGE-NEXT:    pcalau12i $a1, %pc_hi20(__tls_get_addr)
; LA64LARGE-NEXT:    addi.d $ra, $zero, %pc_lo12(__tls_get_addr)
; LA64LARGE-NEXT:    lu32i.d $ra, %pc64_lo20(__tls_get_addr)
; LA64LARGE-NEXT:    lu52i.d $ra, $ra, %pc64_hi12(__tls_get_addr)
; LA64LARGE-NEXT:    add.d $ra, $ra, $a1
; LA64LARGE-NEXT:    jirl $ra, $ra, 0
; LA64LARGE-NEXT:    ld.w $zero, $a0, 0
; LA64LARGE-NEXT:    addi.w $s1, $s1, 1
; LA64LARGE-NEXT:    blt $s1, $fp, .LBB5_1
; LA64LARGE-NEXT:  # %bb.2: # %ret
; LA64LARGE-NEXT:    ld.d $s1, $sp, 0 # 8-byte Folded Reload
; LA64LARGE-NEXT:    ld.d $s0, $sp, 8 # 8-byte Folded Reload
; LA64LARGE-NEXT:    ld.d $fp, $sp, 16 # 8-byte Folded Reload
; LA64LARGE-NEXT:    ld.d $ra, $sp, 24 # 8-byte Folded Reload
; LA64LARGE-NEXT:    addi.d $sp, $sp, 32
; LA64LARGE-NEXT:    ret
entry:
  br label %loop

loop:
  %i = phi i32 [ %inc, %loop ], [ 0, %entry ]
  %0 = load volatile i32, ptr @gd, align 4
  %inc = add nuw nsw i32 %i, 1
  %cmp = icmp slt i32 %inc, %n
  br i1 %cmp, label %loop, label %ret

ret:
  ret void
}