1 files changed, 151 insertions, 0 deletions
diff --git a/llvm/test/CodeGen/X86/callbr-asm-branch-folding.ll b/llvm/test/CodeGen/X86/callbr-asm-branch-folding.ll
new file mode 100644
index 00000000000..bdbff4ac89b
--- /dev/null
+++ b/llvm/test/CodeGen/X86/callbr-asm-branch-folding.ll
@@ -0,0 +1,151 @@
+; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
+; RUN: llc < %s -mtriple=x86_64-unknown-linux-gnu | FileCheck %s
+
+; This test hung in the BranchFolding pass during asm-goto bring up
+
+@e = global i32 0
+@j = global i32 0
+
+define void @n(i32* %o, i32 %p, i32 %u) nounwind {
+; CHECK-LABEL: n:
+; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    pushq %rbp
+; CHECK-NEXT:    pushq %r15
+; CHECK-NEXT:    pushq %r14
+; CHECK-NEXT:    pushq %r13
+; CHECK-NEXT:    pushq %r12
+; CHECK-NEXT:    pushq %rbx
+; CHECK-NEXT:    pushq %rax
+; CHECK-NEXT:    movl %edx, %ebx
+; CHECK-NEXT:    movl %esi, %r12d
+; CHECK-NEXT:    movq %rdi, %r15
+; CHECK-NEXT:    callq c
+; CHECK-NEXT:    movl %eax, %r13d
+; CHECK-NEXT:    movq %r15, %rdi
+; CHECK-NEXT:    callq l
+; CHECK-NEXT:    testl %eax, %eax
+; CHECK-NEXT:    je .LBB0_1
+; CHECK-NEXT:  .LBB0_10: # %cleanup
+; CHECK-NEXT:    addq $8, %rsp
+; CHECK-NEXT:    popq %rbx
+; CHECK-NEXT:    popq %r12
+; CHECK-NEXT:    popq %r13
+; CHECK-NEXT:    popq %r14
+; CHECK-NEXT:    popq %r15
+; CHECK-NEXT:    popq %rbp
+; CHECK-NEXT:    retq
+; CHECK-NEXT:  .LBB0_1: # %if.end
+; CHECK-NEXT:    movl %ebx, {{[-0-9]+}}(%r{{[sb]}}p) # 4-byte Spill
+; CHECK-NEXT:    cmpl $0, {{.*}}(%rip)
+; CHECK-NEXT:    # implicit-def: $ebx
+; CHECK-NEXT:    # implicit-def: $r14d
+; CHECK-NEXT:    je .LBB0_4
+; CHECK-NEXT:  # %bb.2: # %if.then4
+; CHECK-NEXT:    movslq %r12d, %rdi
+; CHECK-NEXT:    callq m
+; CHECK-NEXT:    # implicit-def: $ebx
+; CHECK-NEXT:    # implicit-def: $ebp
+; CHECK-NEXT:  .LBB0_3: # %r
+; CHECK-NEXT:    callq c
+; CHECK-NEXT:    movl %ebp, %r14d
+; CHECK-NEXT:  .LBB0_4: # %if.end8
+; CHECK-NEXT:    movl %ebx, %edi
+; CHECK-NEXT:    callq i
+; CHECK-NEXT:    movl %eax, %ebp
+; CHECK-NEXT:    orl %r14d, %ebp
+; CHECK-NEXT:    testl %r13d, %r13d
+; CHECK-NEXT:    je .LBB0_6
+; CHECK-NEXT:  # %bb.5:
+; CHECK-NEXT:    andl $4, %ebx
+; CHECK-NEXT:    jmp .LBB0_3
+; CHECK-NEXT:  .LBB0_6: # %if.end12
+; CHECK-NEXT:    testl %ebp, %ebp
+; CHECK-NEXT:    je .LBB0_9
+; CHECK-NEXT:  # %bb.7: # %if.then14
+; CHECK-NEXT:    movl {{[-0-9]+}}(%r{{[sb]}}p), %eax # 4-byte Reload
+; CHECK-NEXT:    #APP
+; CHECK-NEXT:    #NO_APP
+; CHECK-NEXT:    jmp .LBB0_10
+; CHECK-NEXT:  .Ltmp0: # Block address taken
+; CHECK-NEXT:  .LBB0_8: # %if.then20.critedge
+; CHECK-NEXT:    movl {{.*}}(%rip), %edi
+; CHECK-NEXT:    movslq %eax, %rcx
+; CHECK-NEXT:    movl $1, %esi
+; CHECK-NEXT:    movq %r15, %rdx
+; CHECK-NEXT:    addq $8, %rsp
+; CHECK-NEXT:    popq %rbx
+; CHECK-NEXT:    popq %r12
+; CHECK-NEXT:    popq %r13
+; CHECK-NEXT:    popq %r14
+; CHECK-NEXT:    popq %r15
+; CHECK-NEXT:    popq %rbp
+; CHECK-NEXT:    jmp k # TAILCALL
+; CHECK-NEXT:  .LBB0_9: # %if.else
+; CHECK-NEXT:    incq 0
+; CHECK-NEXT:    jmp .LBB0_10
+entry:
+  %call = tail call i32 @c()
+  %call1 = tail call i32 @l(i32* %o)
+  %tobool = icmp eq i32 %call1, 0
+  br i1 %tobool, label %if.end, label %cleanup
+
+if.end:                                           ; preds = %entry
+  %0 = load i32, i32* @e
+  %tobool3 = icmp eq i32 %0, 0
+  br i1 %tobool3, label %if.end8, label %if.then4, !prof !0
+
+if.then4:                                         ; preds = %if.end
+  %conv5 = sext i32 %p to i64
+  %call6 = tail call i32 @m(i64 %conv5)
+  br label %r
+
+r:                                                ; preds = %if.end8, %if.then4
+  %flags.0 = phi i32 [ undef, %if.then4 ], [ %and, %if.end8 ]
+  %major.0 = phi i32 [ undef, %if.then4 ], [ %or, %if.end8 ]
+  %call7 = tail call i32 @c()
+  br label %if.end8
+
+if.end8:                                          ; preds = %r, %if.end
+  %flags.1 = phi i32 [ %flags.0, %r ], [ undef, %if.end ]
+  %major.1 = phi i32 [ %major.0, %r ], [ undef, %if.end ]
+  %call9 = tail call i32 @i(i32 %flags.1)
+  %or = or i32 %call9, %major.1
+  %and = and i32 %flags.1, 4
+  %tobool10 = icmp eq i32 %call, 0
+  br i1 %tobool10, label %if.end12, label %r
+
+if.end12:                                         ; preds = %if.end8
+  %tobool13 = icmp eq i32 %or, 0
+  br i1 %tobool13, label %if.else, label %if.then14
+
+if.then14:                                        ; preds = %if.end12
+  callbr void asm sideeffect "", "X,~{dirflag},~{fpsr},~{flags}"(i8* blockaddress(@n, %if.then20.critedge))
+          to label %cleanup [label %if.then20.critedge]
+
+if.then20.critedge:                               ; preds = %if.then14
+  %1 = load i32, i32* @j
+  %conv21 = sext i32 %u to i64
+  %call22 = tail call i32 @k(i32 %1, i64 1, i32* %o, i64 %conv21)
+  br label %cleanup
+
+if.else:                                          ; preds = %if.end12
+  %2 = load i64, i64* null
+  %inc = add i64 %2, 1
+  store i64 %inc, i64* null
+  br label %cleanup
+
+cleanup:                                          ; preds = %if.else, %if.then20.critedge, %if.then14, %entry
+  ret void
+}
+
+declare i32 @c()
+
+declare i32 @l(i32*)
+
+declare i32 @m(i64)
+
+declare i32 @i(i32)
+
+declare i32 @k(i32, i64, i32*, i64)
+
+!0 = !{!"branch_weights", i32 2000, i32 1}