[Bug debug/49828] reversed order of inlined function parameters

["jakub at gcc dot gnu.org" <gcc-bugzilla at gcc dot gnu dot org>]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=49828

Jakub Jelinek <jakub at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |jakub at gcc dot gnu.org

--- Comment #1 from Jakub Jelinek <jakub at gcc dot gnu.org> 2011-07-25 11:02:15 UTC ---
I'd argue that gdb should here use the (correct) order of the parameters on the
abstract origin DIE.
I'm afraid the order of vars everywhere in gcc is pretty much random, and as
the concrete instances of the formal parameters (i.e. vars) can be added
through various means, it is hard to ensure the original ordering.

This particular testcase could be "fixed" by adding
  vars = nreverse (vars);
before:
  declare_inline_vars (id->blocks, vars);
in initialize_inlined_parameters, but that won't help e.g. if a VLA parameter
uses for its size other parameters, or if a function isn't inlined, but cloned
for versioning, etc.