[Bug c++/105569] New: -Waddress warns on dynamic_cast

[Bug c++/105569] New: -Waddress warns on dynamic_cast

[anton.reinhard at protonmail dot com]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=105569

            Bug ID: 105569
           Summary: -Waddress warns on dynamic_cast
           Product: gcc
           Version: 12.1.0
            Status: UNCONFIRMED
          Severity: normal
          Priority: P3
         Component: c++
          Assignee: unassigned at gcc dot gnu.org
          Reporter: anton.reinhard at protonmail dot com
  Target Milestone: ---

compiling with -Waddress yields a useless warning compiling the following code:

Compiler version:
$ gcc -v
Using built-in specs.
COLLECT_GCC=gcc
COLLECT_LTO_WRAPPER=/usr/lib/gcc/x86_64-pc-linux-gnu/12.1.0/lto-wrapper
Target: x86_64-pc-linux-gnu
Configured with: /build/gcc/src/gcc/configure
--enable-languages=c,c++,ada,fortran,go,lto,objc,obj-c++ --enable-bootstrap
--prefix=/usr --libdir=/usr/lib --libexecdir=/usr/lib --mandir=/usr/share/man
--infodir=/usr/share/info --with-bugurl=https://bugs.archlinux.org/
--with-linker-hash-style=gnu --with-system-zlib --enable-__cxa_atexit
--enable-cet=auto --enable-checking=release --enable-clocale=gnu
--enable-default-pie --enable-default-ssp --enable-gnu-indirect-function
--enable-gnu-unique-object --enable-linker-build-id --enable-lto
--enable-multilib --enable-plugin --enable-shared --enable-threads=posix
--disable-libssp --disable-libstdcxx-pch --disable-werror
--with-build-config=bootstrap-lto --enable-link-serialization=1
Thread model: posix
Supported LTO compression algorithms: zlib zstd
gcc version 12.1.0 (GCC) 

Compiled using:
$ g++ warn.cpp -Waddress
warn.cpp: In function ‘int main()’:
warn.cpp:12:22: warning: comparing the result of pointer addition ‘(((A*)ref) +
((sizetype)(*(long int*)((& ref)->B::_vptr.B + -24))))’ and NULL [-Waddress]
   12 |     bool b = nullptr == dynamic_cast<A*>(&ref);
      |              ~~~~~~~~^~~~~~~~~~~~~~~~~~~~~~~~~

The file warn.cpp:
class A {};

class B : public virtual A {};

class C : public A {};

int main() {
    B* object = new B();
    B &ref = *object;

    // -Waddress warns here
    bool b = nullptr == dynamic_cast<A*>(&ref);

    // -Waddress doesn't warn anymore
    auto ptr = dynamic_cast<A*>(&ref); 
    bool b2 = ptr == nullptr;

    C* cobject = new C();
    C &cref = *cobject;

    // -Waddress also doesn't warn anymore
    bool b3 = nullptr == dynamic_cast<A*>(&cref);
}

The warning:
warn.cpp: In function ‘int main()’:
warn.cpp:12:22: warning: comparing the result of pointer addition ‘(((A*)ref) +
((sizetype)(*(long int*)((& ref)->B::_vptr.B + -24))))’ and NULL [-Waddress]
   12 |     bool b = nullptr == dynamic_cast<A*>(&ref);
      |              ~~~~~~~~^~~~~~~~~~~~~~~~~~~~~~~~~

However, dynamic_cast should return nullptr if the cast failed, so this warning
seems invalid to me. I don't know if just the warning is in error or if the
produced code is buggy.
It only happens when the derived class inherits with virtual (see class C)
It doesn't warn in versions previous to gcc-12, or in any clang versions I've
tried which makes me think it's a bug and not me making a mistake.

[Bug c++/105569] [12/13 Regression] -Waddress warns on dynamic_cast

[mpolacek at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=105569

Marek Polacek <mpolacek at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|UNCONFIRMED                 |NEW
            Summary|-Waddress warns on          |[12/13 Regression]
                   |dynamic_cast                |-Waddress warns on
                   |                            |dynamic_cast
                 CC|                            |mpolacek at gcc dot gnu.org,
                   |                            |msebor at gcc dot gnu.org
   Last reconfirmed|                            |2022-05-11
     Ever confirmed|0                           |1
           Priority|P3                          |P2
           Keywords|                            |diagnostic
   Target Milestone|---                         |12.2

--- Comment #1 from Marek Polacek <mpolacek at gcc dot gnu.org> ---
Started with r12-4059-g4dc7ce6fb39179.  Maybe all we need is a
warning_sentinel.

[Bug c++/105569] [12/13 Regression] -Waddress warns on dynamic_cast

[mpolacek at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=105569

Marek Polacek <mpolacek at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |ASSIGNED
           Assignee|unassigned at gcc dot gnu.org      |mpolacek at gcc dot gnu.org

--- Comment #2 from Marek Polacek <mpolacek at gcc dot gnu.org> ---
Candidate fix:

--- a/gcc/cp/rtti.cc
+++ b/gcc/cp/rtti.cc
@@ -844,6 +844,9 @@ build_dynamic_cast (location_t loc, tree type, tree expr,
   if (r != error_mark_node)
     maybe_warn_about_useless_cast (loc, type, expr, complain);
   protected_set_expr_location (r, loc);
+  tree sr = r;
+  STRIP_TYPE_NOPS (sr);
+  suppress_warning (sr, OPT_Waddress);
   return r;
 }

[Bug c++/105569] [12/13 Regression] -Waddress warns on dynamic_cast

[mpolacek at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=105569

--- Comment #3 from Marek Polacek <mpolacek at gcc dot gnu.org> ---
Actually this is probably a better fix; I don't suppose we want to issue
warnings for the compiler-generated POINTER_PLUS_EXPRs.

--- a/gcc/cp/class.cc
+++ b/gcc/cp/class.cc
@@ -518,6 +518,8 @@ build_base_path (enum tree_code code,

   expr = build1 (NOP_EXPR, ptr_target_type, expr);

+  suppress_warning (expr, OPT_Waddress);
+
  indout:
   if (!want_pointer)
     {

[Bug c++/105569] [12/13 Regression] -Waddress warns on dynamic_cast

[cvs-commit at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=105569

--- Comment #4 from CVS Commits <cvs-commit at gcc dot gnu.org> ---
The trunk branch has been updated by Marek Polacek <mpolacek@gcc.gnu.org>:

https://gcc.gnu.org/g:6f56efa94e845db0d5c934ca202295019bf334c1

commit r13-784-g6f56efa94e845db0d5c934ca202295019bf334c1
Author: Marek Polacek <polacek@redhat.com>
Date:   Wed May 11 14:38:49 2022 -0400

    c++: improve -Waddress warnings with *_cast [PR105569]

    This patch improves the diagnostic for -Waddress when it warns for

      if (dynamic_cast<A*>(&ref))
        // ...

    where 'ref' is a reference, which cannot be null.  In particular, it
    changes
    warning: comparing the result of pointer addition '(((A*)ref) +
((sizetype)(*(long int*)((& ref)->B::_vptr.B + -24))))' and NULL
    to
    warning: the compiler can assume that the address of 'ref' will never be
NULL

            PR c++/105569

    gcc/cp/ChangeLog:

            * typeck.cc (warn_for_null_address): Improve the warning when
            the POINTER_PLUS_EXPR's base is of reference type.

    gcc/testsuite/ChangeLog:

            * g++.dg/warn/Waddress-9.C: New test.

[Bug c++/105569] [12/13 Regression] -Waddress warns on dynamic_cast

[mpolacek at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=105569

Marek Polacek <mpolacek at gcc dot gnu.org> changed:

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

--- Comment #5 from Marek Polacek <mpolacek at gcc dot gnu.org> ---
The warning is correct, but I've improved the diagnostic at least.  Not
planning to backport such an improvement.