[Bug c++/98104] New: [11 regression] -Wplacement-new false positive (struture padding?)

[Bug c++/98104] New: [11 regression] -Wplacement-new false positive (struture padding?)

[sss@li-snyder.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=98104

            Bug ID: 98104
           Summary: [11 regression] -Wplacement-new false positive
                    (struture padding?)
           Product: gcc
           Version: 11.0
            Status: UNCONFIRMED
          Severity: normal
          Priority: P3
         Component: c++
          Assignee: unassigned at gcc dot gnu.org
          Reporter: sss@li-snyder.org
  Target Milestone: ---

hi -

With gcc 11 (20201128), the following source (reduced from eigen 3.3.7)
gives a warning:

-- x.cc ---------------------------------------------------------
#include <new>

class Base
{
public:
  Base();
  int* a;
  char b;
};

class Der : public Base {};

void baz()
{
  Der der;
  ::new (static_cast<Base*>(&der)) Base;
}
-----------------------------------------------------------------

$ g++ -c x.cc
x.cc: In function ‘void baz()’:
x.cc:16:10: warning: placement new constructing an object of type ‘Base’ and
size ‘16’ in a region of type ‘Base’ and size ‘9’ [-Wplacement-new=]
   16 |   ::new (static_cast<Base*>(&der)) Base;
      |          ^~~~~~~~~~~~~~~~~~~~~~~~
x.cc:11:7: note: ‘Der::<anonymous>’ declared here
   11 | class Der : public Base {};
      |       ^~~


This appears to be a false positive, possibly due to handling
of padding at the end of the structure.

[Bug c++/98104] [11 regression] -Wplacement-new false positive (struture padding?)

[jakub at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=98104

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> ---
I think this has been fixed with
r11-5628-gb76f83e3859f738809d3aa8bd9dc14e10fc40e24
Might be useful to add the testcase into the testsuite though.

[Bug c++/98104] [11 regression] -Wplacement-new false positive (struture padding?)

[rguenth at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=98104

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

           What    |Removed                     |Added
----------------------------------------------------------------------------
   Target Milestone|---                         |11.0

[Bug c++/98104] [11 regression] -Wplacement-new false positive (struture padding?)

[cvs-commit at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=98104

--- Comment #2 from CVS Commits <cvs-commit at gcc dot gnu.org> ---
The master branch has been updated by Jakub Jelinek <jakub@gcc.gnu.org>:

https://gcc.gnu.org/g:381d1d0120e4dfab3a714f44354e868a80dbaa25

commit r11-5698-g381d1d0120e4dfab3a714f44354e868a80dbaa25
Author: Jakub Jelinek <jakub@redhat.com>
Date:   Thu Dec 3 10:02:02 2020 +0100

    testsuite: Add testcase for already fixed PR [PR98104]

    This testcase got broken with r11-3826 and got fixed with r11-5628.

    2020-12-03  Jakub Jelinek  <jakub@redhat.com>

            PR c++/98104
            * g++.dg/warn/pr98104.C: New test.

[Bug c++/98104] [11 regression] -Wplacement-new false positive (struture padding?)

[jakub at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=98104

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

           What    |Removed                     |Added
----------------------------------------------------------------------------
         Resolution|---                         |FIXED
             Status|UNCONFIRMED                 |RESOLVED

--- Comment #3 from Jakub Jelinek <jakub at gcc dot gnu.org> ---
Testcase installed, the underlying issue is fixed, closing.