operator overloading fails while debugging with gdb for i386.

[Quuxplusone]

Bugzilla Link	PR14354
Status	NEW
Importance	P enhancement
Reported by	Mayur Pandey (mayur.p@samsung.com)
Reported on	2012-11-15 09:51:03 -0800
Last modified on	2014-04-11 21:21:30 -0700
Version	unspecified
Hardware	PC Linux
CC	dblaikie@gmail.com, echristo@gmail.com, llvm-bugs@lists.llvm.org, mayurthebond@gmail.com, rafael@espindo.la
Fixed by commit(s)
Attachments	manual_i386.s (32841 bytes, text/plain)
Blocks
Blocked by
See also

Created attachment 9547
assembly file for the code

For the given test:

class A1 {
  int x;
  int y;

  public:

  A1(int a, int b)
  {
   x=a;
   y=b;
  }

A1 operator+(const A1&);
};

A1 A1::operator+(const A1& second)
{
 A1 sum(0,0);
 sum.x = x + second.x;
 sum.y = y + second.y;

 return (sum);
}

int main (void)
{
 A1 one(2,3);
 A1 two(4,5);

 return 0;
}

when the exectable of this code is debugged in gdb for i386, we dont get the
expected results.

when we break at return 0; and give the command: print one + two, the result
should be $1 = {x = 6, y = 8}, but the result obtained is $1 = {x = 2, y = 3}.

This is related to debug information generated, as normally the overloading is
occuring.
eg: A1 three = one + two results {x = 6, y = 8}.

On checking the assembly, a suspicious entry is found which may be related for
the failure:

    #DEBUG_VALUE: operator+:this <- undef
    #DEBUG_VALUE: operator+:second <- undef

[Quuxplusone]

Attached manual_i386.s (32841 bytes, text/plain): assembly file for the code

[Quuxplusone]

Eric - this was the bug I was mentioning a few days ago.

I haven't looked into/tried to reproduce it yet, but figured one of us will get to it sooner or later

[Quuxplusone]

Was this compiled with -O0 or some optimization?

[Quuxplusone]

it was compiled using -O0.

[Quuxplusone]

In the above case, we see that the operator+ function's return value is
returned in register edx. But gdb expects the return value in eax.

http://stackoverflow.com/questions/267674/inspect-the-return-value-of-a-function-in-gdb

so if we add the assembly instruction: movl     %edx, %eax just before
.loc    1 22 2  instruction in the assembly file and then compile with clang++ and
run the executable, we get the desired results in gdb.

So probably this could be the reason. If this is it we might need to add
another machine instruction. Just needs to be verified properly.

[Quuxplusone]

Still a problem for 32-bit.