[Local] invertCondition: try modifying an existing ICmpInst

[foad]

Summary

This avoids various cases where StructurizeCFG would otherwise insert an
xor i1 instruction, and it since it generally runs late in the pipeline,
instcombine does not clean up the xor-of-cmp pattern.

[foad]

Update remaining AMDGPU tests by hand.

[foad]

foad added a reverting change:

D118623 is an attempt to implement this in StructurizeCFG instead.

llvm/lib/Transforms/Utils/Local.cpp

Value *llvm::invertCondition(Value *Condition) {
   assert(Parent && "Unsupported condition to invert");
 
   // Third: Check all the users for an invert
   for (User *U : Condition->users())
     if (Instruction *I = dyn_cast<Instruction>(U))
       if (I->getParent() == Parent && match(I, m_Not(m_Specific(Condition))))
         return I;
 
+  // Fourth: Modify an existing instruction
+  if (Condition->hasOneUse()) {
+    if (auto *CI = dyn_cast<CmpInst>(Condition)) {
+      CI->setPredicate(CI->getInversePredicate());
+      return Condition;
+    }
+  }
+
   // Last option: Create a new instruction
   auto *Inverted =
       BinaryOperator::CreateNot(Condition, Condition->getName() + ".inv");
   if (Inst && !isa<PHINode>(Inst))
     Inverted->insertAfter(Inst);
   else
     Inverted->insertBefore(&*Parent->getFirstInsertionPt());
   return Inverted;

llvm/test/CodeGen/AMDGPU/GlobalISel/divergent-control-flow.ll

 ; CHECK-NEXT:    s_getpc_b64 s[4:5]
 ; CHECK-NEXT:    s_add_u32 s4, s4, external_constant@gotpcrel32@lo+4
 ; CHECK-NEXT:    s_addc_u32 s5, s5, external_constant@gotpcrel32@hi+12
 ; CHECK-NEXT:    s_load_dwordx2 s[4:5], s[4:5], 0x0
 ; CHECK-NEXT:    s_waitcnt lgkmcnt(0)
 ; CHECK-NEXT:    s_load_dword s4, s[4:5], 0x0
 ; CHECK-NEXT:    s_waitcnt lgkmcnt(0)
 ; CHECK-NEXT:    s_cmp_lg_u32 s4, 0
-; CHECK-NEXT:    s_cbranch_scc1 .LBB4_4
-; CHECK-NEXT:  ; %bb.1: ; %bb2
+; CHECK-NEXT:    s_cbranch_scc0 .LBB4_2
+; CHECK-NEXT:  .LBB4_1: ; %bb12
+; CHECK-NEXT:    s_setpc_b64 s[30:31]
+; CHECK-NEXT:  .LBB4_2: ; %bb2
 ; CHECK-NEXT:    s_getpc_b64 s[4:5]
 ; CHECK-NEXT:    s_add_u32 s4, s4, const.ptr@gotpcrel32@lo+4
 ; CHECK-NEXT:    s_addc_u32 s5, s5, const.ptr@gotpcrel32@hi+12
 ; CHECK-NEXT:    s_load_dwordx2 s[4:5], s[4:5], 0x0
 ; CHECK-NEXT:    v_mov_b32_e32 v0, 0
 ; CHECK-NEXT:    s_waitcnt lgkmcnt(0)
 ; CHECK-NEXT:    s_load_dwordx2 s[4:5], s[4:5], 0x0
 ; CHECK-NEXT:    s_waitcnt lgkmcnt(0)
 ; CHECK-NEXT:    global_load_dword v0, v0, s[4:5]
 ; CHECK-NEXT:    s_mov_b32 s4, -1
 ; CHECK-NEXT:    s_waitcnt vmcnt(0)
 ; CHECK-NEXT:    v_cmp_gt_f32_e32 vcc, 1.0, v0
-; CHECK-NEXT:    s_cbranch_vccnz .LBB4_3
-; CHECK-NEXT:  ; %bb.2: ; %bb7
+; CHECK-NEXT:    s_cbranch_vccnz .LBB4_4
+; CHECK-NEXT:  ; %bb.3: ; %bb7
 ; CHECK-NEXT:    s_mov_b32 s4, 0
-; CHECK-NEXT:  .LBB4_3: ; %bb8
+; CHECK-NEXT:  .LBB4_4: ; %bb8
 ; CHECK-NEXT:    s_cmp_lg_u32 s4, 0
-; CHECK-NEXT:    s_cbranch_scc0 .LBB4_5
-; CHECK-NEXT:  .LBB4_4: ; %bb12
-; CHECK-NEXT:    s_setpc_b64 s[30:31]
-; CHECK-NEXT:  .LBB4_5: ; %bb11
+; CHECK-NEXT:    s_cbranch_scc1 .LBB4_1
+; CHECK-NEXT:  ; %bb.5: ; %bb11
 ; CHECK-NEXT:    v_mov_b32_e32 v0, 4.0
 ; CHECK-NEXT:    buffer_store_dword v0, v0, s[0:3], 0 offen
 ; CHECK-NEXT:    s_waitcnt vmcnt(0)
 ; CHECK-NEXT:    s_setpc_b64 s[30:31]
 bb:
   %tmp = load i32, i32 addrspace(4)* @external_constant
   %ptr = load float*, float* addrspace(4)* @const.ptr
   %tmp1 = icmp ne i32 %tmp, 0

llvm/test/CodeGen/AMDGPU/GlobalISel/sdiv.i64.ll

 ; CGP-NEXT:    v_xor_b32_e32 v3, v5, v0
 ; CGP-NEXT:    v_cndmask_b32_e32 v2, v11, v2, vcc
 ; CGP-NEXT:    v_xor_b32_e32 v0, v1, v3
 ; CGP-NEXT:    v_xor_b32_e32 v1, v2, v3
 ; CGP-NEXT:    v_sub_i32_e32 v0, vcc, v0, v3
 ; CGP-NEXT:    v_subb_u32_e32 v1, vcc, v1, v3, vcc
 ; CGP-NEXT:    ; implicit-def: $vgpr4
 ; CGP-NEXT:    ; implicit-def: $vgpr10
-; CGP-NEXT:  .LBB2_2: ; %Flow2
+; CGP-NEXT:  .LBB2_2: ; %Flow1
 ; CGP-NEXT:    s_or_saveexec_b64 s[6:7], s[6:7]
 ; CGP-NEXT:    s_xor_b64 exec, exec, s[6:7]
 ; CGP-NEXT:    s_cbranch_execz .LBB2_4
 ; CGP-NEXT:  ; %bb.3:
 ; CGP-NEXT:    v_cvt_f32_u32_e32 v0, v4
 ; CGP-NEXT:    v_sub_i32_e32 v1, vcc, 0, v4
 ; CGP-NEXT:    v_rcp_iflag_f32_e32 v0, v0
 ; CGP-NEXT:    v_mul_f32_e32 v0, 0x4f7ffffe, v0

 ; CGP-NEXT:    v_xor_b32_e32 v3, v10, v0
 ; CGP-NEXT:    v_cndmask_b32_e32 v2, v9, v2, vcc
 ; CGP-NEXT:    v_xor_b32_e32 v0, v1, v3
 ; CGP-NEXT:    v_xor_b32_e32 v1, v2, v3
 ; CGP-NEXT:    v_sub_i32_e32 v0, vcc, v0, v3
 ; CGP-NEXT:    v_subb_u32_e32 v1, vcc, v1, v3, vcc
 ; CGP-NEXT:    ; implicit-def: $vgpr2_vgpr3
 ; CGP-NEXT:    ; implicit-def: $vgpr8
-; CGP-NEXT:  .LBB8_2: ; %Flow2
+; CGP-NEXT:  .LBB8_2: ; %Flow1
 ; CGP-NEXT:    s_or_saveexec_b64 s[8:9], s[8:9]
 ; CGP-NEXT:    v_lshl_b64 v[9:10], s[6:7], v6
 ; CGP-NEXT:    s_xor_b64 exec, exec, s[8:9]
 ; CGP-NEXT:    s_cbranch_execz .LBB8_4
 ; CGP-NEXT:  ; %bb.3:
 ; CGP-NEXT:    v_cvt_f32_u32_e32 v0, v2
 ; CGP-NEXT:    v_sub_i32_e32 v1, vcc, 0, v2
 ; CGP-NEXT:    v_rcp_iflag_f32_e32 v0, v0

llvm/test/CodeGen/AMDGPU/GlobalISel/srem.i64.ll

 ; CGP-NEXT:    v_cndmask_b32_e32 v1, v3, v1, vcc
 ; CGP-NEXT:    v_cndmask_b32_e32 v0, v10, v0, vcc
 ; CGP-NEXT:    v_xor_b32_e32 v1, v1, v4
 ; CGP-NEXT:    v_xor_b32_e32 v2, v0, v4
 ; CGP-NEXT:    v_sub_i32_e32 v0, vcc, v1, v4
 ; CGP-NEXT:    v_subb_u32_e32 v1, vcc, v2, v4, vcc
 ; CGP-NEXT:    ; implicit-def: $vgpr4
 ; CGP-NEXT:    ; implicit-def: $vgpr10
-; CGP-NEXT:  .LBB2_2: ; %Flow2
+; CGP-NEXT:  .LBB2_2: ; %Flow1
 ; CGP-NEXT:    s_or_saveexec_b64 s[4:5], s[6:7]
 ; CGP-NEXT:    s_xor_b64 exec, exec, s[4:5]
 ; CGP-NEXT:    s_cbranch_execz .LBB2_4
 ; CGP-NEXT:  ; %bb.3:
 ; CGP-NEXT:    v_cvt_f32_u32_e32 v0, v4
 ; CGP-NEXT:    v_sub_i32_e32 v1, vcc, 0, v4
 ; CGP-NEXT:    v_rcp_iflag_f32_e32 v0, v0
 ; CGP-NEXT:    v_mul_f32_e32 v0, 0x4f7ffffe, v0

 ; CGP-NEXT:    v_cndmask_b32_e32 v1, v3, v1, vcc
 ; CGP-NEXT:    v_cndmask_b32_e32 v0, v9, v0, vcc
 ; CGP-NEXT:    v_xor_b32_e32 v1, v1, v4
 ; CGP-NEXT:    v_xor_b32_e32 v2, v0, v4
 ; CGP-NEXT:    v_sub_i32_e32 v0, vcc, v1, v4
 ; CGP-NEXT:    v_subb_u32_e32 v1, vcc, v2, v4, vcc
 ; CGP-NEXT:    ; implicit-def: $vgpr2_vgpr3
 ; CGP-NEXT:    ; implicit-def: $vgpr8
-; CGP-NEXT:  .LBB8_2: ; %Flow2
+; CGP-NEXT:  .LBB8_2: ; %Flow1
 ; CGP-NEXT:    s_or_saveexec_b64 s[4:5], s[8:9]
 ; CGP-NEXT:    v_lshl_b64 v[9:10], s[6:7], v6
 ; CGP-NEXT:    s_xor_b64 exec, exec, s[4:5]
 ; CGP-NEXT:    s_cbranch_execz .LBB8_4
 ; CGP-NEXT:  ; %bb.3:
 ; CGP-NEXT:    v_cvt_f32_u32_e32 v0, v2
 ; CGP-NEXT:    v_sub_i32_e32 v1, vcc, 0, v2
 ; CGP-NEXT:    v_rcp_iflag_f32_e32 v0, v0

llvm/test/CodeGen/AMDGPU/GlobalISel/udiv.i64.ll

 ; CGP-NEXT:    v_cmp_ne_u32_e32 vcc, 0, v2
 ; CGP-NEXT:    v_cndmask_b32_e32 v2, v14, v12, vcc
 ; CGP-NEXT:    v_cndmask_b32_e32 v3, v15, v16, vcc
 ; CGP-NEXT:    v_cmp_ne_u32_e32 vcc, 0, v10
 ; CGP-NEXT:    v_cndmask_b32_e32 v0, v0, v2, vcc
 ; CGP-NEXT:    v_cndmask_b32_e32 v1, v1, v3, vcc
 ; CGP-NEXT:    ; implicit-def: $vgpr4
 ; CGP-NEXT:    ; implicit-def: $vgpr10
-; CGP-NEXT:  .LBB2_2: ; %Flow2
+; CGP-NEXT:  .LBB2_2: ; %Flow1
 ; CGP-NEXT:    s_or_saveexec_b64 s[6:7], s[6:7]
 ; CGP-NEXT:    s_xor_b64 exec, exec, s[6:7]
 ; CGP-NEXT:    s_cbranch_execz .LBB2_4
 ; CGP-NEXT:  ; %bb.3:
 ; CGP-NEXT:    v_cvt_f32_u32_e32 v0, v4
 ; CGP-NEXT:    v_sub_i32_e32 v1, vcc, 0, v4
 ; CGP-NEXT:    v_rcp_iflag_f32_e32 v0, v0
 ; CGP-NEXT:    v_mul_f32_e32 v0, 0x4f7ffffe, v0

 ; CGP-NEXT:    v_cmp_ne_u32_e32 vcc, 0, v2
 ; CGP-NEXT:    v_cndmask_b32_e32 v2, v13, v11, vcc
 ; CGP-NEXT:    v_cndmask_b32_e32 v3, v14, v15, vcc
 ; CGP-NEXT:    v_cmp_ne_u32_e32 vcc, 0, v9
 ; CGP-NEXT:    v_cndmask_b32_e32 v0, v0, v2, vcc
 ; CGP-NEXT:    v_cndmask_b32_e32 v1, v1, v3, vcc
 ; CGP-NEXT:    ; implicit-def: $vgpr2_vgpr3
 ; CGP-NEXT:    ; implicit-def: $vgpr8
-; CGP-NEXT:  .LBB8_2: ; %Flow2
+; CGP-NEXT:  .LBB8_2: ; %Flow1
 ; CGP-NEXT:    s_or_saveexec_b64 s[8:9], s[8:9]
 ; CGP-NEXT:    v_lshl_b64 v[9:10], s[6:7], v6
 ; CGP-NEXT:    s_xor_b64 exec, exec, s[8:9]
 ; CGP-NEXT:    s_cbranch_execz .LBB8_4
 ; CGP-NEXT:  ; %bb.3:
 ; CGP-NEXT:    v_cvt_f32_u32_e32 v0, v2
 ; CGP-NEXT:    v_sub_i32_e32 v1, vcc, 0, v2
 ; CGP-NEXT:    v_rcp_iflag_f32_e32 v0, v0

llvm/test/CodeGen/AMDGPU/GlobalISel/urem.i64.ll

 ; CGP-NEXT:    v_cmp_ne_u32_e32 vcc, 0, v5
 ; CGP-NEXT:    v_cndmask_b32_e32 v4, v10, v4, vcc
 ; CGP-NEXT:    v_cndmask_b32_e32 v5, v11, v0, vcc
 ; CGP-NEXT:    v_cmp_ne_u32_e32 vcc, 0, v3
 ; CGP-NEXT:    v_cndmask_b32_e32 v0, v1, v4, vcc
 ; CGP-NEXT:    v_cndmask_b32_e32 v1, v2, v5, vcc
 ; CGP-NEXT:    ; implicit-def: $vgpr4
 ; CGP-NEXT:    ; implicit-def: $vgpr10
-; CGP-NEXT:  .LBB2_2: ; %Flow2
+; CGP-NEXT:  .LBB2_2: ; %Flow1
 ; CGP-NEXT:    s_or_saveexec_b64 s[4:5], s[6:7]
 ; CGP-NEXT:    s_xor_b64 exec, exec, s[4:5]
 ; CGP-NEXT:    s_cbranch_execz .LBB2_4
 ; CGP-NEXT:  ; %bb.3:
 ; CGP-NEXT:    v_cvt_f32_u32_e32 v0, v4
 ; CGP-NEXT:    v_sub_i32_e32 v1, vcc, 0, v4
 ; CGP-NEXT:    v_rcp_iflag_f32_e32 v0, v0
 ; CGP-NEXT:    v_mul_f32_e32 v0, 0x4f7ffffe, v0

 ; CGP-NEXT:    v_cmp_ne_u32_e32 vcc, 0, v3
 ; CGP-NEXT:    v_cndmask_b32_e32 v2, v9, v2, vcc
 ; CGP-NEXT:    v_cndmask_b32_e32 v3, v10, v0, vcc
 ; CGP-NEXT:    v_cmp_ne_u32_e32 vcc, 0, v8
 ; CGP-NEXT:    v_cndmask_b32_e32 v0, v1, v2, vcc
 ; CGP-NEXT:    v_cndmask_b32_e32 v1, v4, v3, vcc
 ; CGP-NEXT:    ; implicit-def: $vgpr2_vgpr3
 ; CGP-NEXT:    ; implicit-def: $vgpr8
-; CGP-NEXT:  .LBB8_2: ; %Flow2
+; CGP-NEXT:  .LBB8_2: ; %Flow1
 ; CGP-NEXT:    s_or_saveexec_b64 s[4:5], s[8:9]
 ; CGP-NEXT:    v_lshl_b64 v[9:10], s[6:7], v6
 ; CGP-NEXT:    s_xor_b64 exec, exec, s[4:5]
 ; CGP-NEXT:    s_cbranch_execz .LBB8_4
 ; CGP-NEXT:  ; %bb.3:
 ; CGP-NEXT:    v_cvt_f32_u32_e32 v0, v2
 ; CGP-NEXT:    v_sub_i32_e32 v1, vcc, 0, v2
 ; CGP-NEXT:    v_rcp_iflag_f32_e32 v0, v0

llvm/test/CodeGen/AMDGPU/ctpop16.ll

 define amdgpu_kernel void @ctpop_i16_in_br(i16 addrspace(1)* %out, i16 addrspace(1)* %in, i16 %ctpop_arg, i16 %cond) {
 ; SI-LABEL: ctpop_i16_in_br:
 ; SI:       ; %bb.0: ; %entry
 ; SI-NEXT:    s_load_dword s4, s[0:1], 0xd
 ; SI-NEXT:    s_load_dwordx4 s[0:3], s[0:1], 0x9
 ; SI-NEXT:    s_waitcnt lgkmcnt(0)
 ; SI-NEXT:    s_lshr_b32 s5, s4, 16
 ; SI-NEXT:    s_cmp_lg_u32 s5, 0
-; SI-NEXT:    s_cbranch_scc0 .LBB14_2
+; SI-NEXT:    s_cbranch_scc0 .LBB14_4
 ; SI-NEXT:  ; %bb.1: ; %else
 ; SI-NEXT:    s_mov_b32 s11, 0xf000
 ; SI-NEXT:    s_mov_b32 s10, -1
 ; SI-NEXT:    s_mov_b32 s8, s2
 ; SI-NEXT:    s_mov_b32 s9, s3
 ; SI-NEXT:    buffer_load_ushort v0, off, s[8:11], 0 offset:2
 ; SI-NEXT:    s_mov_b64 s[2:3], 0
-; SI-NEXT:    s_cbranch_execz .LBB14_3
-; SI-NEXT:    s_branch .LBB14_4
-; SI-NEXT:  .LBB14_2:
-; SI-NEXT:    s_mov_b64 s[2:3], -1
-; SI-NEXT:    v_mov_b32_e32 v0, 0
-; SI-NEXT:  .LBB14_3: ; %if
+; SI-NEXT:    s_cbranch_execnz .LBB14_3
+; SI-NEXT:  .LBB14_2: ; %if
 ; SI-NEXT:    s_and_b32 s2, s4, 0xffff
 ; SI-NEXT:    s_bcnt1_i32_b32 s2, s2
 ; SI-NEXT:    s_waitcnt vmcnt(0)
 ; SI-NEXT:    v_mov_b32_e32 v0, s2
-; SI-NEXT:  .LBB14_4: ; %endif
+; SI-NEXT:  .LBB14_3: ; %endif
 ; SI-NEXT:    s_mov_b32 s3, 0xf000
 ; SI-NEXT:    s_mov_b32 s2, -1
 ; SI-NEXT:    s_waitcnt vmcnt(0)
 ; SI-NEXT:    buffer_store_short v0, off, s[0:3], 0
 ; SI-NEXT:    s_endpgm
+; SI-NEXT:  .LBB14_4:
+; SI-NEXT:    s_mov_b64 s[2:3], -1
+; SI-NEXT:    v_mov_b32_e32 v0, 0
+; SI-NEXT:    s_branch .LBB14_2
 ;
 ; VI-LABEL: ctpop_i16_in_br:
 ; VI:       ; %bb.0: ; %entry
 ; VI-NEXT:    s_load_dword s4, s[0:1], 0x34
 ; VI-NEXT:    s_load_dwordx4 s[0:3], s[0:1], 0x24
 ; VI-NEXT:    s_waitcnt lgkmcnt(0)
 ; VI-NEXT:    s_lshr_b32 s5, s4, 16
 ; VI-NEXT:    v_cmp_ne_u16_e64 s[6:7], s5, 0
 ; VI-NEXT:    s_and_b64 vcc, exec, s[6:7]
-; VI-NEXT:    s_cbranch_vccz .LBB14_2
+; VI-NEXT:    s_cbranch_vccz .LBB14_4
 ; VI-NEXT:  ; %bb.1: ; %else
 ; VI-NEXT:    s_mov_b32 s11, 0xf000
 ; VI-NEXT:    s_mov_b32 s10, -1
 ; VI-NEXT:    s_mov_b32 s8, s2
 ; VI-NEXT:    s_mov_b32 s9, s3
 ; VI-NEXT:    buffer_load_ushort v0, off, s[8:11], 0 offset:2
 ; VI-NEXT:    s_mov_b64 s[2:3], 0
-; VI-NEXT:    s_cbranch_execz .LBB14_3
-; VI-NEXT:    s_branch .LBB14_4
-; VI-NEXT:  .LBB14_2:
-; VI-NEXT:    s_mov_b64 s[2:3], -1
-; VI-NEXT:    ; implicit-def: $vgpr0
-; VI-NEXT:  .LBB14_3: ; %if
+; VI-NEXT:    s_cbranch_execnz .LBB14_3
+; VI-NEXT:  .LBB14_2: ; %if
 ; VI-NEXT:    s_and_b32 s2, s4, 0xffff
 ; VI-NEXT:    s_bcnt1_i32_b32 s2, s2
 ; VI-NEXT:    s_waitcnt vmcnt(0)
 ; VI-NEXT:    v_mov_b32_e32 v0, s2
-; VI-NEXT:  .LBB14_4: ; %endif
+; VI-NEXT:  .LBB14_3: ; %endif
 ; VI-NEXT:    s_mov_b32 s3, 0xf000
 ; VI-NEXT:    s_mov_b32 s2, -1
 ; VI-NEXT:    s_waitcnt vmcnt(0)
 ; VI-NEXT:    buffer_store_short v0, off, s[0:3], 0
 ; VI-NEXT:    s_endpgm
+; VI-NEXT:  .LBB14_4:
+; VI-NEXT:    s_mov_b64 s[2:3], -1
+; VI-NEXT:    ; implicit-def: $vgpr0
+; VI-NEXT:    s_branch .LBB14_2
 ;
 ; EG-LABEL: ctpop_i16_in_br:
 ; EG:       ; %bb.0: ; %entry
 ; EG-NEXT:    ALU 0, @20, KC0[], KC1[]
 ; EG-NEXT:    TEX 0 @14
 ; EG-NEXT:    ALU_PUSH_BEFORE 6, @21, KC0[], KC1[]
 ; EG-NEXT:    JUMP @7 POP:1
 ; EG-NEXT:    ALU 0, @28, KC0[CB0:0-32], KC1[]

llvm/test/CodeGen/AMDGPU/insert_vector_elt.ll

 ; the compiler doesn't crash.
 define amdgpu_kernel void @insert_split_bb(<2 x i32> addrspace(1)* %out, i32 addrspace(1)* %in, i32 %a, i32 %b) {
 ; SI-LABEL: insert_split_bb:
 ; SI:       ; %bb.0: ; %entry
 ; SI-NEXT:    s_load_dword s6, s[4:5], 0x4
 ; SI-NEXT:    s_load_dwordx4 s[0:3], s[4:5], 0x0
 ; SI-NEXT:    s_waitcnt lgkmcnt(0)
 ; SI-NEXT:    s_cmp_lg_u32 s6, 0
-; SI-NEXT:    s_cbranch_scc0 .LBB30_2
+; SI-NEXT:    s_cbranch_scc0 .LBB30_4
 ; SI-NEXT:  ; %bb.1: ; %else
 ; SI-NEXT:    s_load_dword s7, s[2:3], 0x1
 ; SI-NEXT:    s_mov_b64 s[4:5], 0
 ; SI-NEXT:    s_andn2_b64 vcc, exec, s[4:5]
 ; SI-NEXT:    s_waitcnt lgkmcnt(0)
 ; SI-NEXT:    s_mov_b64 vcc, vcc
-; SI-NEXT:    s_cbranch_vccz .LBB30_3
-; SI-NEXT:    s_branch .LBB30_4
-; SI-NEXT:  .LBB30_2:
-; SI-NEXT:  .LBB30_3: ; %if
+; SI-NEXT:    s_cbranch_vccnz .LBB30_3
+; SI-NEXT:  .LBB30_2: ; %if
 ; SI-NEXT:    s_load_dword s7, s[2:3], 0x0
-; SI-NEXT:  .LBB30_4: ; %endif
+; SI-NEXT:  .LBB30_3: ; %endif
 ; SI-NEXT:    s_waitcnt lgkmcnt(0)
 ; SI-NEXT:    v_mov_b32_e32 v0, s6
 ; SI-NEXT:    s_mov_b32 s3, 0x100f000
 ; SI-NEXT:    s_mov_b32 s2, -1
 ; SI-NEXT:    v_mov_b32_e32 v1, s7
 ; SI-NEXT:    buffer_store_dwordx2 v[0:1], off, s[0:3], 0
 ; SI-NEXT:    s_endpgm
+; SI-NEXT:  .LBB30_4:
+; SI-NEXT:    s_branch .LBB30_2
 ;
 ; VI-LABEL: insert_split_bb:
 ; VI:       ; %bb.0: ; %entry
 ; VI-NEXT:    s_load_dword s6, s[4:5], 0x10
 ; VI-NEXT:    s_load_dwordx4 s[0:3], s[4:5], 0x0
 ; VI-NEXT:    s_waitcnt lgkmcnt(0)
 ; VI-NEXT:    s_cmp_lg_u32 s6, 0
-; VI-NEXT:    s_cbranch_scc0 .LBB30_2
+; VI-NEXT:    s_cbranch_scc0 .LBB30_4
 ; VI-NEXT:  ; %bb.1: ; %else
 ; VI-NEXT:    s_load_dword s7, s[2:3], 0x4
-; VI-NEXT:    s_cbranch_execz .LBB30_3
-; VI-NEXT:    s_branch .LBB30_4
-; VI-NEXT:  .LBB30_2:
-; VI-NEXT:  .LBB30_3: ; %if
+; VI-NEXT:    s_cbranch_execnz .LBB30_3
+; VI-NEXT:  .LBB30_2: ; %if
 ; VI-NEXT:    s_waitcnt lgkmcnt(0)
 ; VI-NEXT:    s_load_dword s7, s[2:3], 0x0
-; VI-NEXT:  .LBB30_4: ; %endif
+; VI-NEXT:  .LBB30_3: ; %endif
 ; VI-NEXT:    s_waitcnt lgkmcnt(0)
 ; VI-NEXT:    v_mov_b32_e32 v0, s6
 ; VI-NEXT:    s_mov_b32 s3, 0x1100f000
 ; VI-NEXT:    s_mov_b32 s2, -1
 ; VI-NEXT:    v_mov_b32_e32 v1, s7
 ; VI-NEXT:    buffer_store_dwordx2 v[0:1], off, s[0:3], 0
 ; VI-NEXT:    s_endpgm
+; VI-NEXT:  .LBB30_4:
+; VI-NEXT:    s_branch .LBB30_2
 entry:
   %0 = insertelement <2 x i32> undef, i32 %a, i32 0
   %1 = icmp eq i32 %a, 0
   br i1 %1, label %if, label %else
 
 if:
   %2 = load i32, i32 addrspace(1)* %in
   %3 = insertelement <2 x i32> %0, i32 %2, i32 1

llvm/test/CodeGen/AMDGPU/sgpr-control-flow.ll

 define amdgpu_kernel void @sgpr_if_else_salu_br(i32 addrspace(1)* %out, i32 %a, i32 %b, i32 %c, i32 %d, i32 %e) {
 ; SI-LABEL: sgpr_if_else_salu_br:
 ; SI:       ; %bb.0: ; %entry
 ; SI-NEXT:    s_load_dwordx4 s[8:11], s[0:1], 0xb
 ; SI-NEXT:    s_load_dwordx2 s[4:5], s[0:1], 0x9
 ; SI-NEXT:    s_load_dword s0, s[0:1], 0xf
 ; SI-NEXT:    s_waitcnt lgkmcnt(0)
 ; SI-NEXT:    s_cmp_lg_u32 s8, 0
-; SI-NEXT:    s_cbranch_scc0 .LBB0_2
+; SI-NEXT:    s_cbranch_scc0 .LBB0_4
 ; SI-NEXT:  ; %bb.1: ; %else
 ; SI-NEXT:    s_add_i32 s2, s11, s0
-; SI-NEXT:    s_cbranch_execz .LBB0_3
-; SI-NEXT:    s_branch .LBB0_4
-; SI-NEXT:  .LBB0_2:
-; SI-NEXT:    ; implicit-def: $sgpr2
-; SI-NEXT:  .LBB0_3: ; %if
+; SI-NEXT:    s_cbranch_execnz .LBB0_3
+; SI-NEXT:  .LBB0_2: ; %if
 ; SI-NEXT:    s_sub_i32 s2, s9, s10
-; SI-NEXT:  .LBB0_4: ; %endif
+; SI-NEXT:  .LBB0_3: ; %endif
 ; SI-NEXT:    s_add_i32 s0, s2, s8
 ; SI-NEXT:    s_mov_b32 s7, 0xf000
 ; SI-NEXT:    s_mov_b32 s6, -1
 ; SI-NEXT:    v_mov_b32_e32 v0, s0
 ; SI-NEXT:    buffer_store_dword v0, off, s[4:7], 0
 ; SI-NEXT:    s_endpgm
+; SI-NEXT:  .LBB0_4:
+; SI-NEXT:    ; implicit-def: $sgpr2
+; SI-NEXT:    s_branch .LBB0_2
 
 entry:
   %0 = icmp eq i32 %a, 0
   br i1 %0, label %if, label %else
 
 if:
   %1 = sub i32 %b, %c
   br label %endif

 
 define amdgpu_kernel void @sgpr_if_else_salu_br_opt(i32 addrspace(1)* %out, [8 x i32], i32 %a, [8 x i32], i32 %b, [8 x i32], i32 %c, [8 x i32], i32 %d, [8 x i32], i32 %e) {
 ; SI-LABEL: sgpr_if_else_salu_br_opt:
 ; SI:       ; %bb.0: ; %entry
 ; SI-NEXT:    s_load_dword s6, s[0:1], 0x13
 ; SI-NEXT:    s_load_dwordx2 s[4:5], s[0:1], 0x9
 ; SI-NEXT:    s_waitcnt lgkmcnt(0)
 ; SI-NEXT:    s_cmp_lg_u32 s6, 0
-; SI-NEXT:    s_cbranch_scc0 .LBB1_2
+; SI-NEXT:    s_cbranch_scc0 .LBB1_4
 ; SI-NEXT:  ; %bb.1: ; %else
 ; SI-NEXT:    s_load_dword s2, s[0:1], 0x2e
 ; SI-NEXT:    s_load_dword s3, s[0:1], 0x37
 ; SI-NEXT:    s_waitcnt lgkmcnt(0)
 ; SI-NEXT:    s_add_i32 s7, s2, s3
-; SI-NEXT:    s_cbranch_execz .LBB1_3
-; SI-NEXT:    s_branch .LBB1_4
-; SI-NEXT:  .LBB1_2:
-; SI-NEXT:    ; implicit-def: $sgpr7
-; SI-NEXT:  .LBB1_3: ; %if
+; SI-NEXT:    s_cbranch_execnz .LBB1_3
+; SI-NEXT:  .LBB1_2: ; %if
 ; SI-NEXT:    s_load_dword s2, s[0:1], 0x1c
 ; SI-NEXT:    s_load_dword s0, s[0:1], 0x25
 ; SI-NEXT:    s_waitcnt lgkmcnt(0)
 ; SI-NEXT:    s_add_i32 s7, s2, s0
-; SI-NEXT:  .LBB1_4: ; %endif
+; SI-NEXT:  .LBB1_3: ; %endif
 ; SI-NEXT:    s_add_i32 s0, s7, s6
 ; SI-NEXT:    s_mov_b32 s7, 0xf000
 ; SI-NEXT:    s_mov_b32 s6, -1
 ; SI-NEXT:    v_mov_b32_e32 v0, s0
 ; SI-NEXT:    buffer_store_dword v0, off, s[4:7], 0
 ; SI-NEXT:    s_endpgm
+; SI-NEXT:  .LBB1_4:
+; SI-NEXT:    ; implicit-def: $sgpr7
+; SI-NEXT:    s_branch .LBB1_2
 
 entry:
   %cmp0 = icmp eq i32 %a, 0
   br i1 %cmp0, label %if, label %else
 
 if:
   %add0 = add i32 %b, %c
   br label %endif

llvm/test/Transforms/StructurizeCFG/AMDGPU/loop-subregion-misordered.ll

 ; BB1 and BB3. This fact will lead BB2 to be placed in the wrong order.
 ;
 ; In this work, we treat the SubRegion as a special case and use its exit block to determine
 ; the loop and its depth to guard the sorting.
 define amdgpu_kernel void @loop_subregion_misordered(i32 addrspace(1)* %arg0) #0 {
 ; CHECK-LABEL: @loop_subregion_misordered(
 ; CHECK-NEXT:  entry:
 ; CHECK-NEXT:    [[TMP:%.*]] = load volatile <2 x i32>, <2 x i32> addrspace(1)* undef, align 16
-; CHECK-NEXT:    [[LOAD1:%.*]] = load volatile <2 x float>, <2 x float> addrspace(1)* undef
+; CHECK-NEXT:    [[LOAD1:%.*]] = load volatile <2 x float>, <2 x float> addrspace(1)* undef, align 8
 ; CHECK-NEXT:    [[TID:%.*]] = call i32 @llvm.amdgcn.workitem.id.x()
 ; CHECK-NEXT:    [[GEP:%.*]] = getelementptr inbounds i32, i32 addrspace(1)* [[ARG0:%.*]], i32 [[TID]]
 ; CHECK-NEXT:    [[I_INITIAL:%.*]] = load volatile i32, i32 addrspace(1)* [[GEP]], align 4
 ; CHECK-NEXT:    br label [[LOOP_HEADER:%.*]]
 ; CHECK:       LOOP.HEADER:
-; CHECK-NEXT:    [[I:%.*]] = phi i32 [ [[I_INITIAL]], [[ENTRY:%.*]] ], [ [[TMP4:%.*]], [[FLOW3:%.*]] ]
+; CHECK-NEXT:    [[I:%.*]] = phi i32 [ [[I_INITIAL]], [[ENTRY:%.*]] ], [ [[TMP3:%.*]], [[FLOW3:%.*]] ]
 ; CHECK-NEXT:    call void asm sideeffect "s_nop 0x100b
 ; CHECK-NEXT:    [[TMP12:%.*]] = zext i32 [[I]] to i64
 ; CHECK-NEXT:    [[TMP13:%.*]] = getelementptr inbounds <4 x i32>, <4 x i32> addrspace(1)* null, i64 [[TMP12]]
 ; CHECK-NEXT:    [[TMP14:%.*]] = load <4 x i32>, <4 x i32> addrspace(1)* [[TMP13]], align 16
 ; CHECK-NEXT:    [[TMP15:%.*]] = extractelement <4 x i32> [[TMP14]], i64 0
 ; CHECK-NEXT:    [[TMP16:%.*]] = and i32 [[TMP15]], 65535
-; CHECK-NEXT:    [[TMP17:%.*]] = icmp eq i32 [[TMP16]], 1
-; CHECK-NEXT:    [[TMP0:%.*]] = xor i1 [[TMP17]], true
-; CHECK-NEXT:    br i1 [[TMP0]], label [[BB62:%.*]], label [[FLOW:%.*]]
+; CHECK-NEXT:    [[TMP17:%.*]] = icmp ne i32 [[TMP16]], 1
+; CHECK-NEXT:    br i1 [[TMP17]], label [[BB62:%.*]], label [[FLOW:%.*]]
 ; CHECK:       Flow1:
-; CHECK-NEXT:    [[TMP1:%.*]] = phi i32 [ [[INC_I:%.*]], [[INCREMENT_I:%.*]] ], [ undef, [[BB62]] ]
-; CHECK-NEXT:    [[TMP2:%.*]] = phi i1 [ false, [[INCREMENT_I]] ], [ true, [[BB62]] ]
-; CHECK-NEXT:    [[TMP3:%.*]] = phi i1 [ true, [[INCREMENT_I]] ], [ false, [[BB62]] ]
+; CHECK-NEXT:    [[TMP0:%.*]] = phi i32 [ [[INC_I:%.*]], [[INCREMENT_I:%.*]] ], [ undef, [[BB62]] ]
+; CHECK-NEXT:    [[TMP1:%.*]] = phi i1 [ false, [[INCREMENT_I]] ], [ true, [[BB62]] ]
+; CHECK-NEXT:    [[TMP2:%.*]] = phi i1 [ true, [[INCREMENT_I]] ], [ false, [[BB62]] ]
 ; CHECK-NEXT:    br label [[FLOW]]
 ; CHECK:       bb18:
 ; CHECK-NEXT:    [[TMP19:%.*]] = extractelement <2 x i32> [[TMP]], i64 0
 ; CHECK-NEXT:    [[TMP22:%.*]] = lshr i32 [[TMP19]], 16
 ; CHECK-NEXT:    [[TMP24:%.*]] = urem i32 [[TMP22]], 52
 ; CHECK-NEXT:    [[TMP25:%.*]] = mul nuw nsw i32 [[TMP24]], 52
 ; CHECK-NEXT:    br label [[INNER_LOOP:%.*]]
 ; CHECK:       Flow2:
-; CHECK-NEXT:    [[TMP4]] = phi i32 [ [[TMP59:%.*]], [[INNER_LOOP_BREAK:%.*]] ], [ [[TMP8:%.*]], [[FLOW]] ]
-; CHECK-NEXT:    [[TMP5:%.*]] = phi i1 [ true, [[INNER_LOOP_BREAK]] ], [ [[TMP10:%.*]], [[FLOW]] ]
-; CHECK-NEXT:    br i1 [[TMP5]], label [[END_ELSE_BLOCK:%.*]], label [[FLOW3]]
+; CHECK-NEXT:    [[TMP3]] = phi i32 [ [[TMP59:%.*]], [[INNER_LOOP_BREAK:%.*]] ], [ [[TMP6:%.*]], [[FLOW]] ]
+; CHECK-NEXT:    [[TMP4:%.*]] = phi i1 [ true, [[INNER_LOOP_BREAK]] ], [ [[TMP8:%.*]], [[FLOW]] ]
+; CHECK-NEXT:    br i1 [[TMP4]], label [[END_ELSE_BLOCK:%.*]], label [[FLOW3]]
 ; CHECK:       INNER_LOOP:
 ; CHECK-NEXT:    [[INNER_LOOP_J:%.*]] = phi i32 [ [[INNER_LOOP_J_INC:%.*]], [[INNER_LOOP]] ], [ [[TMP25]], [[BB18:%.*]] ]
 ; CHECK-NEXT:    call void asm sideeffect "
 ; CHECK-NEXT:    [[INNER_LOOP_J_INC]] = add nsw i32 [[INNER_LOOP_J]], 1
 ; CHECK-NEXT:    [[INNER_LOOP_CMP:%.*]] = icmp eq i32 [[INNER_LOOP_J]], 0
 ; CHECK-NEXT:    br i1 [[INNER_LOOP_CMP]], label [[INNER_LOOP_BREAK]], label [[INNER_LOOP]]
 ; CHECK:       INNER_LOOP_BREAK:
 ; CHECK-NEXT:    [[TMP59]] = extractelement <4 x i32> [[TMP14]], i64 2
-; CHECK-NEXT:    call void asm sideeffect "s_nop 23 ", "~{memory}"() #0
+; CHECK-NEXT:    call void asm sideeffect "s_nop 23 ", "~{memory}"() #[[ATTR0:[0-9]+]]
 ; CHECK-NEXT:    br label [[FLOW2:%.*]]
 ; CHECK:       bb62:
-; CHECK-NEXT:    [[LOAD13:%.*]] = icmp ult i32 [[TMP16]], 271
-; CHECK-NEXT:    [[TMP6:%.*]] = xor i1 [[LOAD13]], true
-; CHECK-NEXT:    br i1 [[TMP6]], label [[INCREMENT_I]], label [[FLOW1:%.*]]
+; CHECK-NEXT:    [[LOAD13:%.*]] = icmp uge i32 [[TMP16]], 271
+; CHECK-NEXT:    br i1 [[LOAD13]], label [[INCREMENT_I]], label [[FLOW1:%.*]]
 ; CHECK:       Flow3:
-; CHECK-NEXT:    [[TMP7:%.*]] = phi i1 [ [[CMP_END_ELSE_BLOCK:%.*]], [[END_ELSE_BLOCK]] ], [ true, [[FLOW2]] ]
-; CHECK-NEXT:    br i1 [[TMP7]], label [[FLOW4:%.*]], label [[LOOP_HEADER]]
+; CHECK-NEXT:    [[TMP5:%.*]] = phi i1 [ [[CMP_END_ELSE_BLOCK:%.*]], [[END_ELSE_BLOCK]] ], [ true, [[FLOW2]] ]
+; CHECK-NEXT:    br i1 [[TMP5]], label [[FLOW4:%.*]], label [[LOOP_HEADER]]
 ; CHECK:       Flow4:
-; CHECK-NEXT:    br i1 [[TMP9:%.*]], label [[BB64:%.*]], label [[RETURN:%.*]]
+; CHECK-NEXT:    br i1 [[TMP7:%.*]], label [[BB64:%.*]], label [[RETURN:%.*]]
 ; CHECK:       bb64:
-; CHECK-NEXT:    call void asm sideeffect "s_nop 42", "~{memory}"() #0
+; CHECK-NEXT:    call void asm sideeffect "s_nop 42", "~{memory}"() #[[ATTR0]]
 ; CHECK-NEXT:    br label [[RETURN]]
 ; CHECK:       Flow:
-; CHECK-NEXT:    [[TMP8]] = phi i32 [ [[TMP1]], [[FLOW1]] ], [ undef, [[LOOP_HEADER]] ]
-; CHECK-NEXT:    [[TMP9]] = phi i1 [ [[TMP2]], [[FLOW1]] ], [ false, [[LOOP_HEADER]] ]
-; CHECK-NEXT:    [[TMP10]] = phi i1 [ [[TMP3]], [[FLOW1]] ], [ false, [[LOOP_HEADER]] ]
-; CHECK-NEXT:    [[TMP11:%.*]] = phi i1 [ false, [[FLOW1]] ], [ true, [[LOOP_HEADER]] ]
-; CHECK-NEXT:    br i1 [[TMP11]], label [[BB18]], label [[FLOW2]]
+; CHECK-NEXT:    [[TMP6]] = phi i32 [ [[TMP0]], [[FLOW1]] ], [ undef, [[LOOP_HEADER]] ]
+; CHECK-NEXT:    [[TMP7]] = phi i1 [ [[TMP1]], [[FLOW1]] ], [ false, [[LOOP_HEADER]] ]
+; CHECK-NEXT:    [[TMP8]] = phi i1 [ [[TMP2]], [[FLOW1]] ], [ false, [[LOOP_HEADER]] ]
+; CHECK-NEXT:    [[TMP9:%.*]] = phi i1 [ false, [[FLOW1]] ], [ true, [[LOOP_HEADER]] ]
+; CHECK-NEXT:    br i1 [[TMP9]], label [[BB18]], label [[FLOW2]]
 ; CHECK:       INCREMENT_I:
 ; CHECK-NEXT:    [[INC_I]] = add i32 [[I]], 1
 ; CHECK-NEXT:    call void asm sideeffect "s_nop 0x1336
 ; CHECK-NEXT:    br label [[FLOW1]]
 ; CHECK:       END_ELSE_BLOCK:
 ; CHECK-NEXT:    call void asm sideeffect "s_nop 0x1337
-; CHECK-NEXT:    [[CMP_END_ELSE_BLOCK]] = icmp eq i32 [[TMP4]], -1
+; CHECK-NEXT:    [[CMP_END_ELSE_BLOCK]] = icmp eq i32 [[TMP3]], -1
 ; CHECK-NEXT:    br label [[FLOW3]]
 ; CHECK:       RETURN:
 ; CHECK-NEXT:    call void asm sideeffect "s_nop 0x99
 ; CHECK-NEXT:    store volatile <2 x float> [[LOAD1]], <2 x float> addrspace(1)* undef, align 8
 ; CHECK-NEXT:    ret void
 ;
 entry:
   %tmp = load volatile <2 x i32>, <2 x i32> addrspace(1)* undef, align 16

llvm/test/Transforms/StructurizeCFG/bug36015.ll

 
 loop.inner:
   %ctr.loop.inner = phi i32 [ %ctr.loop.outer, %loop.outer ], [ %ctr.if, %if ]
   call void @foo(i32 1)
   %cond.inner = icmp eq i32 %cmp0, %ctr.loop.inner
   br i1 %cond.inner, label %if, label %else
 
 ; CHECK: if:
-; CHECK:   %cond.if.inv = xor i1 %cond.if, true
+; CHECK:   %cond.if = icmp sge i32 %ctr.if, %count
 ; CHECK:   br label %Flow
 if:
   %ctr.if = add i32 %ctr.loop.inner, 1
   call void @foo(i32 2)
   %cond.if = icmp slt i32 %ctr.if, %count
   br i1 %cond.if, label %loop.inner, label %exit
 
 ; CHECK: Flow:
-; CHECK:   %1 = phi i1 [ %cond.if.inv, %if ], [ true, %loop.inner ]
+; CHECK:   %1 = phi i1 [ %cond.if, %if ], [ true, %loop.inner ]
 ; CHECK:   %2 = phi i1 [ false, %if ], [ true, %loop.inner ]
 ; CHECK:   br i1 %1, label %Flow1, label %loop.inner
 
 ; CHECK: Flow1:
 ; CHECK:   br i1 %2, label %else, label %Flow2
 
 ; CHECK: else:
 ; CHECK:   br label %Flow2
 else:
   %ctr.else = add i32 %ctr.loop.inner, 1
   call void @foo(i32 3)
   %cond.else = icmp slt i32 %ctr.else, %count
   br i1 %cond.else, label %loop.outer, label %exit
 
 ; CHECK: Flow2:
-; CHECK:   %4 = phi i1 [ %cond.else.inv, %else ], [ true, %Flow1 ]
+; CHECK:   %4 = phi i1 [ %cond.else, %else ], [ true, %Flow1 ]
 ; CHECK:   br i1 %4, label %exit, label %loop.outer
 
 exit:
   ret void
 }
 
 declare void @foo(i32)

llvm/test/Transforms/StructurizeCFG/invert-constantexpr.ll

 ; NOTE: Assertions have been autogenerated by utils/update_test_checks.py
 ; RUN: opt -S -o - -structurizecfg < %s | FileCheck %s
 
 @g = global i32 0
 
 define void @invert_constantexpr_condition(i32 %arg, i32 %arg1) #0 {
 ; CHECK-LABEL: @invert_constantexpr_condition(
 ; CHECK-NEXT:  bb:
-; CHECK-NEXT:    [[TMP:%.*]] = icmp eq i32 [[ARG:%.*]], 0
-; CHECK-NEXT:    [[TMP0:%.*]] = xor i1 [[TMP]], true
+; CHECK-NEXT:    [[TMP:%.*]] = icmp ne i32 [[ARG:%.*]], 0
 ; CHECK-NEXT:    br i1 icmp eq (i32 ptrtoint (i32* @g to i32), i32 0), label [[BB2:%.*]], label [[FLOW:%.*]]
 ; CHECK:       bb2:
 ; CHECK-NEXT:    br label [[FLOW]]
 ; CHECK:       bb3:
 ; CHECK-NEXT:    [[TMP4:%.*]] = phi i1 [ undef, [[FLOW]] ], [ [[TMP7:%.*]], [[BB6:%.*]] ]
 ; CHECK-NEXT:    [[TMP5:%.*]] = or i1 [[TMP4]], icmp eq (i32 ptrtoint (i32* @g to i32), i32 0)
 ; CHECK-NEXT:    br label [[BB8:%.*]]
 ; CHECK:       Flow:
-; CHECK-NEXT:    [[TMP1:%.*]] = phi i1 [ [[TMP0]], [[BB2]] ], [ icmp ne (i32 ptrtoint (i32* @g to i32), i32 0), [[BB:%.*]] ]
-; CHECK-NEXT:    br i1 [[TMP1]], label [[BB6]], label [[BB3:%.*]]
+; CHECK-NEXT:    [[TMP0:%.*]] = phi i1 [ [[TMP]], [[BB2]] ], [ icmp ne (i32 ptrtoint (i32* @g to i32), i32 0), [[BB:%.*]] ]
+; CHECK-NEXT:    br i1 [[TMP0]], label [[BB6]], label [[BB3:%.*]]
 ; CHECK:       bb6:
 ; CHECK-NEXT:    [[TMP7]] = icmp slt i32 [[ARG]], [[ARG1:%.*]]
 ; CHECK-NEXT:    br label [[BB3]]
 ; CHECK:       bb8:
 ; CHECK-NEXT:    ret void
 ;
 bb:
   %tmp = icmp eq i32 %arg, 0

llvm/test/Transforms/StructurizeCFG/one-loop-multiple-backedges.ll

 ; RUN: opt -S -structurizecfg %s -o - | FileCheck %s
 
 ; CHECK-NOT: br i1 true
 
 define void @blam(i32 addrspace(1)* nocapture %arg, float %arg1, float %arg2) {
 ; CHECK: bb:
 bb:
   br label %bb3
 
 ; CHECK: bb3:
-; CHECK:   %tmp4.inv = xor i1 %tmp4, true
-; CHECK:   br i1 %tmp4.inv, label %bb5, label %Flow
+; CHECK:   %tmp4 = fcmp oge float %arg1, 3.500000e+00
+; CHECK:   br i1 %tmp4, label %bb5, label %Flow
 bb3:                                              ; preds = %bb7, %bb
   %tmp = phi i64 [ 0, %bb ], [ %tmp8, %bb7 ]
   %tmp4 = fcmp ult float %arg1, 3.500000e+00
   br i1 %tmp4, label %bb7, label %bb5
 
 ; CHECK: bb5:
-; CHECK:   %tmp6.inv = xor i1 %tmp6, true
+; CHECK:   %tmp6 = fcmp uge float 0.000000e+00, %arg2
 ; CHECK:   br label %Flow
 bb5:                                              ; preds = %bb3
   %tmp6 = fcmp olt float 0.000000e+00, %arg2
   br i1 %tmp6, label %bb10, label %bb7
 
 ; CHECK: Flow:
-; CHECK:   %0 = phi i1 [ %tmp6.inv, %bb5 ], [ %tmp4, %bb3 ]
+; CHECK:   %0 = phi i1 [ %tmp6, %bb5 ], [ %tmp4, %bb3 ]
 ; CHECK:   br i1 %0, label %bb7, label %Flow1
 
 ; CHECK: bb7:
 ; CHECK:   br label %Flow1
 bb7:                                              ; preds = %bb5, %bb3
   %tmp8 = add nuw nsw i64 %tmp, 1
   %tmp9 = icmp slt i64 %tmp8, 5
   br i1 %tmp9, label %bb3, label %bb10
 
 ; CHECK: Flow1:
-; CHECK:   %3 = phi i1 [ %tmp9.inv, %bb7 ], [ true, %Flow ]
+; CHECK:   %3 = phi i1 [ %tmp9, %bb7 ], [ true, %Flow ]
 ; CHECK:   br i1 %3, label %bb10, label %bb3
 
 ; CHECK: bb10:
 bb10:                                             ; preds = %bb7, %bb5
   %tmp11 = phi i32 [ 15, %bb5 ], [ 255, %bb7 ]
   store i32 %tmp11, i32 addrspace(1)* %arg, align 4
   ret void
 }

llvm/test/Transforms/StructurizeCFG/post-order-traversal-bug.ll

for.end:                                          ; preds = %for.body.1, %if.then
   %best_val.0.lcssa = phi float [ %best_val.233, %for.body.1 ], [ %tmp5, %if.then ]
   store float %best_val.0.lcssa, float* %out, align 4
   ret void
 
 ; CHECK: Flow1
 ; CHECK: br i1 %{{[0-9]}}, label %for.body.1, label %Flow2
 
 ; CHECK: for.body.1:
-; CHECK: br i1 %cmp1.5.inv, label %for.body.6, label %Flow3
+; CHECK: br i1 %cmp1.5, label %for.body.6, label %Flow3
 for.body.1:                                       ; preds = %if.then, %lor.lhs.false
   %best_val.233 = phi float [ %tmp5, %if.then ], [ %best_val.027, %lor.lhs.false ]
   %best_count.231 = phi i32 [ %sub4, %if.then ], [ %best_count.025, %lor.lhs.false ]
   %indvars.iv.next.454 = add nsw i64 %indvars.iv, 5
   %tmp22 = trunc i64 %indvars.iv.next.454 to i32
   %cmp1.5 = icmp eq i32 %tmp22, %K1
   br i1 %cmp1.5, label %for.end, label %for.body.6