[LLVMbugs] [Bug 5176] New: small testcase shows much badness (EH, constant folding)

[bugzilla-daemon at cs.uiuc.edu]

http://llvm.org/bugs/show_bug.cgi?id=5176

           Summary: small testcase shows much badness (EH, constant folding)
           Product: tools
           Version: 2.2
          Platform: PC
        OS/Version: All
            Status: NEW
          Severity: normal
          Priority: P2
         Component: llvm-g++
        AssignedTo: unassignedbugs at nondot.org
        ReportedBy: clattner at apple.com
                CC: llvmbugs at cs.uiuc.edu


Consider this testcase:
struct A
{
  bool foo(int*) const;
} a;

struct B {};

struct B1 : B
{
  bool (A::*pmf)(int*) const;
  const A* pa;

  B1() : pmf(&A::foo), pa(&a) {}
  bool operator()() const { return (pa->*pmf)(new int); }
};

struct B2 : B
{
  B1 b1;

  B2(const B1& _b1) : b1(_b1) {}
  bool operator()() const { return b1(); }
};

template<int> struct C
{
  void bar(B2 b2) { while (b2()) ; }
  C() { bar(B2(B1())); }
};

void baz(int i){
  switch(i) {
    case 0: new C<0>;
    case 1: new C<1>;
    case 2: new C<2>;
  }
}

When compiled with:

$ llvm-gcc t.cc -S -o - -emit-llvm -O3

There are two problems.  First, we get a bunch of horrible unfolded constant
expressions like this:

inttoptr (i64 trunc (i256 lshr (i256 or (i256 and (i256 and (i256 shl (i256
zext (i64 ptrtoint (i8 (%struct.A*, i32*)* @_ZNK1A3fooEPi to i64) to i256),
i256 64), i256 -6277101735386680763495507056286727952638980837032266301441),
i256 6277101735386680763835789423207666416102355444464034512895), i256 shl
(i256 zext (i64 ptrtoint (%struct.A* @a to i64) to i256), i256 192)), i256 64)
to i64) to i8 (%struct.A*, i32*)*)

Quite impressive, but not very useful.

The second problem is EH related, we get 3 copies of the exact same landing
pad:

pad:                                             ; preds = %.noexc,
%_ZNK2B2clEv.exit.i.i
  %eh_ptr = tail call i8* @llvm.eh.exception()    ; <i8*> [#uses=2]
  %eh_select14 = tail call i64 (i8*, i8*, ...)* @llvm.eh.selector.i64(i8*
%eh_ptr, i8* bitcast (i32 (...)* @__gxx_personality_v0 to i8*), i8* null) ;
<i64> [#uses=0]
  tail call void @_ZdlPv(i8* %0) nounwind
  br label %Unwind

lpad15:                                           ; preds = %.noexc41,
%_ZNK2B2clEv.exit.i.i40
  %eh_ptr16 = tail call i8* @llvm.eh.exception()  ; <i8*> [#uses=2]
  %eh_select18 = tail call i64 (i8*, i8*, ...)* @llvm.eh.selector.i64(i8*
%eh_ptr16, i8* bitcast (i32 (...)* @__gxx_personality_v0 to i8*), i8* null) ;
<i64> [#uses=0]
  tail call void @_ZdlPv(i8* %8) nounwind
  br label %Unwind

lpad19:                                           ; preds = %.noexc33,
%_ZNK2B2clEv.exit.i.i32
  %eh_ptr20 = tail call i8* @llvm.eh.exception()  ; <i8*> [#uses=2]
  %eh_select22 = tail call i64 (i8*, i8*, ...)* @llvm.eh.selector.i64(i8*
%eh_ptr20, i8* bitcast (i32 (...)* @__gxx_personality_v0 to i8*), i8* null) ;
<i64> [#uses=0]
  tail call void @_ZdlPv(i8* %16) nounwind
  br label %Unwind

-Chris


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.