[Bug debug/94450] lto abstract variable emitted as concrete decl

[rguenth at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=94450

--- Comment #3 from Richard Biener <rguenth at gcc dot gnu.org> ---
(In reply to Richard Biener from comment #2)
> So the current situation is similar to that of

Modifying the testcase to C99

inline int foo(int i)
{
  static int j;
  j = i + 1;
  return j;
}

int bar(int i)
{
  return foo(i);
}

int baz(int i)
{
  return foo(i);
}

extern int foo(int i);

[...]

> Note we're not emitting a concrete instance of the function which would
> contain the "main" instance of 'j'.  Not sure what the DWARF for a
> standalone concrete instantiation of 'j' would look like.

The concrete instance is emitted:

 <1><57>: Abbrev Number: 6 (DW_TAG_subprogram)
    <58>   DW_AT_abstract_origin: <0x2d>
    <5c>   DW_AT_low_pc      : 0x0
    <64>   DW_AT_high_pc     : 0xa
    <6c>   DW_AT_frame_base  : 1 byte block: 9c         (DW_OP_call_frame_cfa)
    <6e>   DW_AT_GNU_all_call_sites: 1
    <6e>   DW_AT_sibling     : <0x89>
 <2><72>: Abbrev Number: 7 (DW_TAG_formal_parameter)
    <73>   DW_AT_abstract_origin: <0x3d>
    <77>   DW_AT_location    : 1 byte block: 55         (DW_OP_reg5 (rdi))
 <2><79>: Abbrev Number: 8 (DW_TAG_variable)
    <7a>   DW_AT_abstract_origin: <0x46>
    <7e>   DW_AT_location    : 9 byte block: 3 0 0 0 0 0 0 0 0  (DW_OP_addr: 0)

so here's another copy of the variable.

I guess for function local variables gdb isn't fooled to think it has
two copies?  And the bogus copy is not because of the abstract origin
link but because of the unit import but could be brought in by other
means of making gdb read the DWARF of it?