[Bug c++/94510] New: nullptr_t implicitly cast to zero twice in std::array

[Bug c++/94510] New: nullptr_t implicitly cast to zero twice in std::array

[kndevl at outlook dot com]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=94510

            Bug ID: 94510
           Summary: nullptr_t implicitly cast to zero twice in std::array
           Product: gcc
           Version: 9.3.0
            Status: UNCONFIRMED
          Severity: normal
          Priority: P3
         Component: c++
          Assignee: unassigned at gcc dot gnu.org
          Reporter: kndevl at outlook dot com
  Target Milestone: ---

`std::array<int, 3> arr{ nullptr, 0, 0 };` is expected to fail to compile
similar to how `std::array<int, 3> arr{ nullptr, 0, 1 };` fails. What I infer
is that if all elements are effectively zero, type conversions are not checked.
This makes sense if a variable of static storage duration is initialized to
zero and std::array constructor can ignore the arguments in the initializer if
all arguments are implicitly zero. 

Is there anything on the C++ standard that allows this optimization I am
missing here?

# This works fine, as expected

`g++ -save-temps -c not-bug.cpp`

> not-bug.cpp: In function ‘void test()’:
> not-bug.cpp:4:43: error: cannot convert ‘std::nullptr_t’ to ‘int’ in initialization
>     4 |     std::array<int, 3> arr{ nullptr, 0, 1 };

# This compiles fine, but this should error out

`g++ -save-temps -c bug.cpp`


I built the tip of gcc yesterday. It had the same bug.
Any pointers on how I can go about debugging this? 
- Does type checking happen during GENERIC stage?
- How do I print the source string corresponding to a TREE struct?

[Bug c++/94510] nullptr_t implicitly cast to zero twice in std::array

[kndevl at outlook dot com]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=94510

--- Comment #1 from Karthik Nishanth <kndevl at outlook dot com> ---
Created attachment 48218
  --> https://gcc.gnu.org/bugzilla/attachment.cgi?id=48218&action=edit
preprocessed source for bug.cpp

[Bug c++/94510] nullptr_t implicitly cast to zero twice in std::array

[kndevl at outlook dot com]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=94510

--- Comment #2 from Karthik Nishanth <kndevl at outlook dot com> ---
Created attachment 48219
  --> https://gcc.gnu.org/bugzilla/attachment.cgi?id=48219&action=edit
bug.cpp

[Bug c++/94510] nullptr_t implicitly cast to zero twice in std::array

[kndevl at outlook dot com]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=94510

--- Comment #3 from Karthik Nishanth <kndevl at outlook dot com> ---
Created attachment 48220
  --> https://gcc.gnu.org/bugzilla/attachment.cgi?id=48220&action=edit
complete command line invocations

[Bug c++/94510] nullptr_t implicitly cast to zero twice in std::array

[kndevl at outlook dot com]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=94510

--- Comment #4 from Karthik Nishanth <kndevl at outlook dot com> ---
Created attachment 48221
  --> https://gcc.gnu.org/bugzilla/attachment.cgi?id=48221&action=edit
gcc version string

[Bug c++/94510] nullptr_t implicitly cast to zero twice in std::array

[kndevl at outlook dot com]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=94510

--- Comment #5 from Karthik Nishanth <kndevl at outlook dot com> ---
Created attachment 48222
  --> https://gcc.gnu.org/bugzilla/attachment.cgi?id=48222&action=edit
machine type

[Bug c++/94510] nullptr_t implicitly cast to zero twice in std::array

[kndevl at outlook dot com]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=94510

--- Comment #6 from Karthik Nishanth <kndevl at outlook dot com> ---
Created attachment 48223
  --> https://gcc.gnu.org/bugzilla/attachment.cgi?id=48223&action=edit
not a bug

[Bug c++/94510] nullptr_t implicitly cast to zero twice in std::array

[kndevl at outlook dot com]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=94510

--- Comment #7 from Karthik Nishanth <kndevl at outlook dot com> ---
Created attachment 48224
  --> https://gcc.gnu.org/bugzilla/attachment.cgi?id=48224&action=edit
not a bug, preprocessed

[Bug c++/94510] [9/10 Regression] nullptr_t implicitly cast to zero twice in std::array

[redi at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=94510

Jonathan Wakely <redi at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
      Known to fail|                            |10.0, 9.3.0
      Known to work|                            |8.4.0
   Last reconfirmed|                            |2020-04-07
            Summary|nullptr_t implicitly cast   |[9/10 Regression] nullptr_t
                   |to zero twice in std::array |implicitly cast to zero
                   |                            |twice in std::array
     Ever confirmed|0                           |1
   Target Milestone|---                         |9.4
             Status|UNCONFIRMED                 |NEW
                 CC|                            |msebor at gcc dot gnu.org

--- Comment #8 from Jonathan Wakely <redi at gcc dot gnu.org> ---
(In reply to Karthik Nishanth from comment #4)
> Created attachment 48221 [details]
> gcc version string

There's no need to create attachments for each tiny bit of text, comments are
fine, and much easier for us to see, without having to click through a separate
page. (Your attachment in comment 3 just duplicates what you already said in
comment 0 anyway!)

Relevant info from the comment 4 attachment:

Using built-in specs.
COLLECT_GCC=g++
COLLECT_LTO_WRAPPER=/usr/lib/gcc/x86_64-pc-linux-gnu/9.3.0/lto-wrapper
Target: x86_64-pc-linux-gnu
Configured with: /build/gcc/src/gcc/configure --prefix=/usr --libdir=/usr/lib
--libexecdir=/usr/lib --mandir=/usr/share/man --infodir=/usr/share/info
--with-pkgversion='Arch Linux 9.3.0-1'
--with-bugurl=https://bugs.archlinux.org/
--enable-languages=c,c++,ada,fortran,go,lto,objc,obj-c++,d --enable-shared
--enable-threads=posix --with-system-zlib --with-isl --enable-__cxa_atexit
--disable-libunwind-exceptions --enable-clocale=gnu --disable-libstdcxx-pch
--disable-libssp --enable-gnu-unique-object --enable-linker-build-id
--enable-lto --enable-plugin --enable-install-libiberty
--with-linker-hash-style=gnu --enable-gnu-indirect-function --enable-multilib
--disable-werror --enable-checking=release --enable-default-pie
--enable-default-ssp --enable-cet=auto gdc_include_dir=/usr/include/dlang/gdc
Thread model: posix
gcc version 9.3.0 (Arch Linux 9.3.0-1) 


Reduced testcase:

int arr[3] = { nullptr, 0, 0 };

This is a regression introduced by r270155

    PR c++/89974 - ICE on a definition of a non-type specialization on a struct
object with pointer to member function

    PR c++/89974 - ICE on a definition of a non-type specialization on a struct
object with pointer to member function
    PR c++/89878 - same specializations on a zero-initialized struct object as
a non-type parameter treated as distinct
    PR c++/89833 - sorry, unimplemented: string literal in function template
signature
    PR c++/47488 - sorry, unimplemented: string literal in function template
signature

    gcc/cp/ChangeLog:

            PR c++/89974
            PR c++/89878
            PR c++/89833
            PR c++/47488
            * decl.c (reshape_init_array_1): Strip trailing zero-initializers
            from arrays of trivial type and known size.
            * mangle.c (write_expression): Convert braced initializer lists
            to STRING_CSTs.
            (write_expression): Trim trailing zero-initializers from arrays
            of trivial type.
            (write_template_arg_literal): Mangle strings the same as braced
            initializer lists.

    gcc/testsuite/ChangeLog:

            PR c++/89974
            PR c++/89878
            PR c++/89833
            PR c++/47488
            * gcc/testsuite/g++.dg/abi/mangle69.C: New test.
            * gcc/testsuite/g++.dg/abi/mangle70.C: New test.
            * gcc/testsuite/g++.dg/abi/mangle71.C: New test.
            * gcc/testsuite/g++.dg/abi/mangle72.C: New test.
            * gcc/testsuite/g++.dg/cpp0x/constexpr-array19.C: New test.
            * gcc/testsuite/g++.dg/cpp2a/nontype-class15.C: New test.
            * gcc/testsuite/g++.dg/cpp2a/nontype-class16.C: New test.
            * gcc/testsuite/g++.dg/init/array51.C: New test.

[Bug c++/94510] [9/10 Regression] nullptr_t implicitly cast to zero twice in std::array

[msebor at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=94510

Martin Sebor <msebor at gcc dot gnu.org> changed:

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

--- Comment #9 from Martin Sebor <msebor at gcc dot gnu.org> ---
The changed code in reshape_init_array_1 checks for integer/pointer mismatches
via the following test:

      /* Pointers initialized to strings must be treated as non-zero
         even if the string is empty.  */
      tree init_type = TREE_TYPE (elt_init);
      if (POINTER_TYPE_P (elt_type) != POINTER_TYPE_P (init_type)
          || !type_initializer_zero_p (elt_type, elt_init))
        last_nonzero = index;

...and a test case involving an ordinary pointer is diagnosed:

$ cat t.C && gcc -S -Wall t.C
int a[2] = { nullptr };
int b[2] = { (void*)0 };

t.C:2:23: error: invalid conversion from ‘void*’ to ‘int’ [-fpermissive]
    2 | int b[2] = { (void*)0 };
      |                       ^
      |                       |
      |                       void*


but a nullptr use is not.

(gdb) p debug_tree (init_type)
$6 = (tree) 0x7fffea9521f8
<nullptr_type 0x7fffea9521f8 decltype(nullptr) public unsigned DI
    size <integer_cst 0x7fffea7f7e70 type <integer_type 0x7fffea8150a8
bitsizetype> constant 64>
    unit-size <integer_cst 0x7fffea7f7e88 type <integer_type 0x7fffea815000
sizetype> constant 8>
    align:64 warn_if_not_align:0 symtab:0 alias-set -1 canonical-type
0x7fffea9521f8>

I would expect POINTER_TYPE_P (init_type) to be non-zero but it's not:

(gdb) p POINTER_TYPE_P (init_type)
$7 = false

[Bug c++/94510] [9/10 Regression] nullptr_t implicitly cast to zero twice in std::array

[msebor at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=94510

Martin Sebor <msebor at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
           Keywords|                            |patch

--- Comment #10 from Martin Sebor <msebor at gcc dot gnu.org> ---
POINTER_TYPE_P() is zero because while nullptr is specified to be "the pointer
literal is the keyword nullptr" it also "...is a prvalue of type
std::nullptr_t. [Note: std::nullptr_t is a distinct type that is neither a
pointer type nor a pointer-to-member type; ..."  Of course -- a pointer literal
that doesn't have a pointer type, makes perfect sense.

Another closely related regression r270155 introduced is the missing warning
for the following test case:

$ cat t.C && gcc-8 -S -Wall t.C
#include <stddef.h>

int a[2] = { NULL };

t.C:3:19: warning: converting to non-pointer type ‘int’ from NULL
[-Wconversion-null]
 int a[2] = { NULL };
                   ^

The patch at the link below restores both diagnostics:
  https://gcc.gnu.org/pipermail/gcc-patches/2020-April/543482.html

[Bug c++/94510] [9/10 Regression] nullptr_t implicitly cast to zero twice in std::array

[cvs-commit at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=94510

--- Comment #11 from CVS Commits <cvs-commit at gcc dot gnu.org> ---
The master branch has been updated by Jason Merrill <jason@gcc.gnu.org>:

https://gcc.gnu.org/g:587970215f4681def390e2a791aa3ba6adb65158

commit r10-7864-g587970215f4681def390e2a791aa3ba6adb65158
Author: Martin Sebor <msebor@gmail.com>
Date:   Wed Apr 22 02:27:54 2020 -0400

    c++: reject scalar array initialization with nullptr [PR94510]

    The change committed to GCC 9 to allow string literals as template
arguments
    caused the compiler to prune away, and thus miss diagnosing, conversion
from
    nullptr to int in an array initializer.  After looking at various
approaches
    to improving the pruning, we realized that the only place the pruning is
    necessary is in the mangler.

    gcc/cp/ChangeLog
    2020-04-22  Martin Sebor  <msebor@redhat.com>
                Jason Merrill  <jason@redhat.com>

            PR c++/94510
            * decl.c (reshape_init_array_1): Avoid stripping redundant trailing
            zero initializers...
            * mangle.c (write_expression): ...and handle them here even for
            pointers to members by calling zero_init_expr_p.
            * cp-tree.h (zero_init_expr_p): Declare.
            * tree.c (zero_init_expr_p): Define.
            (type_initializer_zero_p): Remove.
            * pt.c (tparm_obj_values): New hash_map.
            (get_template_parm_object): Store to it.
            (tparm_object_argument): New.

    gcc/testsuite/ChangeLog
    2020-04-22  Martin Sebor  <msebor@redhat.com>

            PR c++/94510
            * g++.dg/init/array58.C: New test.
            * g++.dg/init/array59.C: New test.
            * g++.dg/cpp2a/nontype-class34.C: New test.
            * g++.dg/cpp2a/nontype-class35.C: New test.

[Bug c++/94510] [9 Regression] nullptr_t implicitly cast to zero twice in std::array

[jason at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=94510

Jason Merrill <jason at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |jason at gcc dot gnu.org
            Summary|[9/10 Regression] nullptr_t |[9 Regression] nullptr_t
                   |implicitly cast to zero     |implicitly cast to zero
                   |twice in std::array         |twice in std::array

--- Comment #12 from Jason Merrill <jason at gcc dot gnu.org> ---
Fixed for GCC 10 so far.

[Bug c++/94510] [9 Regression] nullptr_t implicitly cast to zero twice in std::array

[cvs-commit at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=94510

--- Comment #13 from CVS Commits <cvs-commit at gcc dot gnu.org> ---
The releases/gcc-9 branch has been updated by Jason Merrill
<jason@gcc.gnu.org>:

https://gcc.gnu.org/g:35699e75a502bb25a2cb23a456e8d6f036b8f226

commit r9-8526-g35699e75a502bb25a2cb23a456e8d6f036b8f226
Author: Martin Sebor <msebor@gmail.com>
Date:   Tue Apr 21 11:02:06 2020 -0400

    c++: reject scalar array initialization with nullptr [PR94510]

    The change committed to GCC 9 to allow string literals as template
arguments
    caused the compiler to prune away, and thus miss diagnosing, conversion
from
    nullptr to int in an array initializer.  After looking at various
approaches
    to improving the pruning, we realized that the only place the pruning is
    necessary is in the mangler.

    gcc/cp/ChangeLog
    2020-04-21  Martin Sebor  <msebor@redhat.com>
                Jason Merrill  <jason@redhat.com>

            PR c++/94510
            * decl.c (reshape_init_array_1): Avoid stripping redundant trailing
            zero initializers...
            * mangle.c (write_expression): ...and handle them here even for
            pointers to members by calling zero_init_expr_p.
            * cp-tree.h (zero_init_expr_p): Declare.
            * tree.c (zero_init_expr_p): Define.
            (type_initializer_zero_p): Remove.
            * pt.c (tparm_obj_values): New hash_map.
            (get_template_parm_object): Store to it.
            (tparm_object_argument): New.

    gcc/testsuite/ChangeLog
    2020-04-21  Martin Sebor  <msebor@redhat.com>

            PR c++/94510
            * g++.dg/init/array58.C: New test.
            * g++.dg/init/array59.C: New test.
            * g++.dg/cpp2a/nontype-class34.C: New test.
            * g++.dg/cpp2a/nontype-class35.C: New test.

[Bug c++/94510] [9 Regression] nullptr_t implicitly cast to zero twice in std::array

[jason at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=94510

Jason Merrill <jason at gcc dot gnu.org> changed:

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

--- Comment #14 from Jason Merrill <jason at gcc dot gnu.org> ---
Fixed for 9.4/10.