[Bug debug/60815] New: Inconsistent prologue line table location

[Bug debug/60815] New: Inconsistent prologue line table location

[dblaikie at gmail dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=60815

            Bug ID: 60815
           Summary: Inconsistent prologue line table location
           Product: gcc
           Version: 4.9.0
            Status: UNCONFIRMED
          Severity: normal
          Priority: P3
         Component: debug
          Assignee: unassigned at gcc dot gnu.org
          Reporter: dblaikie at gmail dot com
                CC: ccoutant at gcc dot gnu.org, echristo at gmail dot com
              Host: x86_64
            Target: x86_64

Given:

template<typename T>
void func() // prologue
{
}

template void func<int>();

void f2()
{ // prologue
}

struct foo {
  void f3() // GCC prologue
  {
  }
  void f4();
};

void (foo::*x)() = &foo::f3;

void foo::f4()
{ // prologue
}

GCC's line table shows the prologue of both of these functions as the line of
the function name, not the opening '{'. Yet if these functions are
non-templates or out-of-line, the prologue starts at the opening brace, not the
function name.

(backstory: Clang consistently uses the opening brace but fails some GDB tests
that rely on GCC's behavior, even though it's inconsistent (see
gdb.cp/cpexprs.exp for some examples). It'd be great if both GCC and Clang
could agree on this, one way or another, but if not, the GDB tests can of
course be made resilient to the difference by simply putting the '{' on the
same line as the function name)

[Bug debug/60815] Inconsistent prologue line table location

[dblaikie at gmail dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=60815

--- Comment #1 from David Blaikie <dblaikie at gmail dot com> ---
Oh - and if we can confirm the direction you're going with this (if the
decision is that the prologue should start, like Clang, at the opening '{'
always, for example) I'll go ahead and update the GDB test suite to either be
agnostic or to KFAIL those tests under GDB, referencing this bug.

[Bug c++/60815] Inconsistent prologue line table location

[rguenth at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=60815

Richard Biener <rguenth at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
           Keywords|                            |wrong-debug
          Component|debug                       |c++

--- Comment #2 from Richard Biener <rguenth at gcc dot gnu.org> ---
Probably cfun->function_start_locus is inconsistent for them (or not set, in
which case we use DECL_SOURCE_LOCATION).

Sounds like a C++ frontend issue to me:

static tree
cp_parser_init_declarator (cp_parser* parser,
                           cp_decl_specifier_seq *decl_specifiers,
                           vec<deferred_access_check, va_gc> *checks,
                           bool function_definition_allowed_p,
                           bool member_p,
                           int declares_class_or_enum,
                           bool* function_definition_p,
                           tree* maybe_range_for_decl)
{
...
          if (decl != error_mark_node && DECL_STRUCT_FUNCTION (decl))
            {
              /* This is where the prologue starts...  */
              DECL_STRUCT_FUNCTION (decl)->function_start_locus
                = func_brace_location;

but for templates we likely don't have DECL_STRUCT_FUNCTION allocated at this
point.

[Bug c++/60815] prologue line is wrong for templates

[pinskia at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=60815

Andrew Pinski <pinskia at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|UNCONFIRMED                 |NEW
            Summary|Inconsistent prologue line  |prologue line is wrong for
                   |table location              |templates
     Ever confirmed|0                           |1
   Last reconfirmed|                            |2021-08-27

--- Comment #4 from Andrew Pinski <pinskia at gcc dot gnu.org> ---
Confirmed,
For templates, we should save off func_brace_location (somewhere) and set
DECL_STRUCT_FUNCTION (decl)->function_start_locus when the function gets
instantiated.