[Bug c++/94476] New: NSDMI deferred parse

[Bug c++/94476] New: NSDMI deferred parse

[nathan at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=94476

            Bug ID: 94476
           Summary: NSDMI deferred parse
           Product: gcc
           Version: 10.0
            Status: UNCONFIRMED
          Severity: normal
          Priority: P3
         Component: c++
          Assignee: unassigned at gcc dot gnu.org
          Reporter: nathan at gcc dot gnu.org
  Target Milestone: ---

This code, extracted from <filesystem> leaves an unparsed NSDMI for _M_type. 
That kills the module streamer, which doesn't expect to meet such things. 
Bizarrely, removing either the 'signed char' underlying type, the ctor
declaration, or the _M_path field removes the problem.  Not sure if the
non-module compiler copes?

enum class file_type : signed char { none = 0 };

class directory_entry
{
public:
  directory_entry(int);

  int _M_path;

  file_type _M_type = file_type::none;
};

[Bug c++/94476] NSDMI deferred parse

[nathan at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=94476

Nathan Sidwell <nathan at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|UNCONFIRMED                 |ASSIGNED
   Last reconfirmed|                            |2020-04-03
     Ever confirmed|0                           |1
           Assignee|unassigned at gcc dot gnu.org      |nathan at gcc dot gnu.org

[Bug c++/94476] NSDMI deferred parse

[nathan at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=94476

Nathan Sidwell <nathan at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
           Priority|P3                          |P4
             Status|ASSIGNED                    |SUSPENDED

--- Comment #1 from Nathan Sidwell <nathan at gcc dot gnu.org> ---
this caused by the as_base creation code copying the field_decl, deferred NSDMI
and all.  That doesn't affect trunk's behaviour so is harmless.  But we should
get it right.  Fixed on modules 2c6b4bd9ab0

Can push to trunk once stage 1 opens up.

        PR c++/94476
        gcc/cp/
        * class.c (layout_class_type): Zap NSDMI of as_base fields.
        gcc/testsuite/
        * g++.dg/modules/nsdmi-2.C: New.

[Bug c++/94476] NSDMI deferred parse

[ppalka at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=94476

Patrick Palka <ppalka at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |ppalka at gcc dot gnu.org
   Target Milestone|---                         |11.0
         Resolution|---                         |FIXED
             Status|SUSPENDED                   |RESOLVED

--- Comment #2 from Patrick Palka <ppalka at gcc dot gnu.org> ---
Fixed since r11-2371-gb95eba48a1a252 I suppose.