[Bug c++/107079] New: ICE initializing lifetime-extended constexpr variable that stores its this pointer

[Bug c++/107079] New: ICE initializing lifetime-extended constexpr variable that stores its this pointer

[johelegp at gmail dot com]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=107079

            Bug ID: 107079
           Summary: ICE initializing lifetime-extended constexpr variable
                    that stores its this pointer
           Product: gcc
           Version: 13.0
            Status: UNCONFIRMED
          Severity: normal
          Priority: P3
         Component: c++
          Assignee: unassigned at gcc dot gnu.org
          Reporter: johelegp at gmail dot com
                CC: johelegp at gmail dot com
  Target Milestone: ---

See https://godbolt.org/z/5asqcnoMe.

```C++
struct X {
  const X* x = this;
};
constexpr const X& x = X{};
```

```
<source>:4:26: internal compiler error: in cxx_eval_constant_expression, at
cp/constexpr.cc:7591
    4 | constexpr const X& x = X{};
      |                          ^
0x23507de internal_error(char const*, ...)
        ???:0
0xa9ea08 fancy_abort(char const*, int, char const*)
        ???:0
0xd4f2da store_init_value(tree_node*, tree_node*, vec<tree_node*, va_gc,
vl_embed>**, int)
        ???:0
0xb78f38 cp_finish_decl(tree_node*, tree_node*, bool, tree_node*, int)
        ???:0
0xc786f7 c_parse_file()
        ???:0
0xdb4a19 c_common_parse_file()
        ???:0
Please submit a full bug report, with preprocessed source (by using
-freport-bug).
Please include the complete backtrace with any bug report.
See <https://gcc.gnu.org/bugs/> for instructions.
Compiler returned: 1
```

[Bug c++/107079] [10/11/12/13 Regression] ICE initializing lifetime-extended constexpr variable that stores its this pointer

[mpolacek at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=107079

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

           What    |Removed                     |Added
----------------------------------------------------------------------------
   Last reconfirmed|                            |2022-09-29
            Summary|ICE initializing            |[10/11/12/13 Regression]
                   |lifetime-extended constexpr |ICE initializing
                   |variable that stores its    |lifetime-extended constexpr
                   |this pointer                |variable that stores its
                   |                            |this pointer
   Target Milestone|---                         |10.5
     Ever confirmed|0                           |1
             Status|UNCONFIRMED                 |NEW
                 CC|                            |jason at gcc dot gnu.org,
                   |                            |mpolacek at gcc dot gnu.org
           Priority|P3                          |P2

--- Comment #1 from Marek Polacek <mpolacek at gcc dot gnu.org> ---
Confirmed.  Started with r269003:

commit e8b3c1bc3ba22dcf59b9c743f11d4cb2bc5d7792
Author: Jason Merrill <jason@redhat.com>
Date:   Mon Feb 18 20:01:50 2019 -0500

    PR c++/89336 - multiple stores in constexpr stmt.

[Bug c++/107079] [10/11/12/13 Regression] ICE initializing lifetime-extended constexpr variable that stores its this pointer

[mpolacek at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=107079

--- Comment #2 from Marek Polacek <mpolacek at gcc dot gnu.org> ---
Very interesting.  We're in store_init_value, initializing x with
&TARGET_EXPR <D.2768, {.x=(const struct X *) &<PLACEHOLDER_EXPR struct X>}>
but we must lifetime-extend via extend_ref_init_temps and we get

_ZGR1x_.x = (const struct X *) &<PLACEHOLDER_EXPR struct X> >>>;, (const struct
X &) &_ZGR1x_;

Since x was declared constexpr, we do cxx_constant_init (not quiet) and we hit
the code added in r269003:

+  bool preeval = SCALAR_TYPE_P (type) || TREE_CODE (t) == MODIFY_EXPR;
+  if (preeval)
+    {
+      /* Evaluate the value to be stored without knowing what object it will
be
+    stored in, so that any side-effects happen first.  */
+      if (!SCALAR_TYPE_P (type))
+   new_ctx.ctor = new_ctx.object = NULL_TREE;
+      init = cxx_eval_constant_expression (&new_ctx, init, false,
+                      non_constant_p, overflow_p);

while evaluating the INIT_EXPR _ZGR1x_.x = (const struct X *)
&<PLACEHOLDER_EXPR struct X> >>>.  But we have no ctx.ctor or ctx.object here
so lookup_placeholder won't find anything and we ICE on
 7861       /* A placeholder without a referent.  We can get here when
 7862          checking whether NSDMIs are noexcept, or in massage_init_elt;
 7863          just say it's non-constant for now.  */
 7864       gcc_assert (ctx->quiet);

Somehow we should manage that the PLACEHOLDER_EXPR is replaced with _ZGR1x_ I
guess.

[Bug c++/107079] [10/11/12/13 Regression] ICE initializing lifetime-extended constexpr variable that stores its this pointer

[mpolacek at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=107079

--- Comment #3 from Marek Polacek <mpolacek at gcc dot gnu.org> ---
The cxx_constant_init call actually takes decl=x so we should probably use
that.
value = cxx_constant_init (value, decl);

However, in cxx_eval_outermost_constant_expr type is const struct X & and so we
don't set up ctx.object.

[Bug c++/107079] [10/11/12/13 Regression] ICE initializing lifetime-extended constexpr variable that stores its this pointer

[mpolacek at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=107079

--- Comment #4 from Marek Polacek <mpolacek at gcc dot gnu.org> ---
The ICE could be fixed with

--- a/gcc/cp/call.cc
+++ b/gcc/cp/call.cc
@@ -13604,9 +13604,13 @@ set_up_extended_ref_temp (tree decl, tree expr,
vec<tree, va_gc> **cleanups,
       init = NULL_TREE;
     }
   else
-    /* Create the INIT_EXPR that will initialize the temporary
-       variable.  */
-    init = split_nonconstant_init (var, expr);
+    {
+      /* Create the INIT_EXPR that will initialize the temporary
+    variable.  */
+      init = split_nonconstant_init (var, expr);
+      replace_placeholders (init, var);
+    }
+
   if (at_function_scope_p ())
     {
       add_decl_expr (var);


but it'll result in
error: modification of ‘<temporary>’ is not a constant expression
because the INIT_EXPR is
_ZGR1x_.x = (const struct X *) &_ZGR1x_
but _ZGR1x_ isn't in our constexpr context.

[Bug c++/107079] [10/11/12/13 Regression] ICE initializing lifetime-extended constexpr variable that stores its this pointer

[mpolacek at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=107079

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 #5 from Marek Polacek <mpolacek at gcc dot gnu.org> ---
I'll post...something.

[Bug c++/107079] [10/11/12/13 Regression] ICE initializing lifetime-extended constexpr variable that stores its this pointer

[cvs-commit at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=107079

--- Comment #6 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:67b82bc1b29b82e4577df9491793624f3a8eb12f

commit r13-5763-g67b82bc1b29b82e4577df9491793624f3a8eb12f
Author: Marek Polacek <polacek@redhat.com>
Date:   Wed Feb 8 14:02:48 2023 -0500

    c++: ICE initing lifetime-extended constexpr var [PR107079]

    We ICE on the simple:

      struct X { const X* x = this; };
      constexpr const X& x = X{};

    where store_init_value initializes 'x' with

      &TARGET_EXPR <D.2768, {.x=(const struct X *) &<PLACEHOLDER_EXPR struct
X>}>

    but we must lifetime-extend via extend_ref_init_temps and we get

      _ZGR1x_.x = (const struct X *) &<PLACEHOLDER_EXPR struct X> >>>;, (const
struct X &) &_ZGR1x_;

    Since 'x' was declared constexpr, we do cxx_constant_init and we hit
    the preeval code added in r269003 while evaluating the INIT_EXPR:

      _ZGR1x_.x = (const struct X *) &<PLACEHOLDER_EXPR struct X> >>>

    but we have no ctx.ctor or ctx.object here so lookup_placeholder won't
    find anything that could replace X and we ICE on
     7861       /* A placeholder without a referent.  We can get here when
     7862          checking whether NSDMIs are noexcept, or in
massage_init_elt;
     7863          just say it's non-constant for now.  */
     7864       gcc_assert (ctx->quiet);
    because cxx_constant_init means !ctx->quiet.  It's not correct that
    there isn't a referent.  I think the following patch is a pretty
    straightforward fix: pass the _ZGR var down to maybe_constant_init so
    that it can replace the PLACEHOLDER_EXPR with _ZGR1x_.

    The commented assert in the test doesn't pass: we complain that _ZGR1x_
    isn't a constexpr variable because we don't implement DR2126 (PR101588).

            PR c++/107079

    gcc/cp/ChangeLog:

            * call.cc (set_up_extended_ref_temp): Pass var to
maybe_constant_init.

    gcc/testsuite/ChangeLog:

            * g++.dg/cpp0x/constexpr-nsdmi2.C: New test.

[Bug c++/107079] [10/11/12 Regression] ICE initializing lifetime-extended constexpr variable that stores its this pointer

[mpolacek at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=107079

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

           What    |Removed                     |Added
----------------------------------------------------------------------------
            Summary|[10/11/12/13 Regression]    |[10/11/12 Regression] ICE
                   |ICE initializing            |initializing
                   |lifetime-extended constexpr |lifetime-extended constexpr
                   |variable that stores its    |variable that stores its
                   |this pointer                |this pointer

--- Comment #7 from Marek Polacek <mpolacek at gcc dot gnu.org> ---
Fixed on trunk so far.

[Bug c++/107079] [10/11/12 Regression] ICE initializing lifetime-extended constexpr variable that stores its this pointer

[cvs-commit at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=107079

--- Comment #8 from CVS Commits <cvs-commit at gcc dot gnu.org> ---
The releases/gcc-12 branch has been updated by Marek Polacek
<mpolacek@gcc.gnu.org>:

https://gcc.gnu.org/g:fb82334341e21ad0254f63e942be276f62d111cf

commit r12-9233-gfb82334341e21ad0254f63e942be276f62d111cf
Author: Marek Polacek <polacek@redhat.com>
Date:   Wed Feb 8 14:02:48 2023 -0500

    c++: ICE initing lifetime-extended constexpr var [PR107079]

    We ICE on the simple:

      struct X { const X* x = this; };
      constexpr const X& x = X{};

    where store_init_value initializes 'x' with

      &TARGET_EXPR <D.2768, {.x=(const struct X *) &<PLACEHOLDER_EXPR struct
X>}>

    but we must lifetime-extend via extend_ref_init_temps and we get

      _ZGR1x_.x = (const struct X *) &<PLACEHOLDER_EXPR struct X> >>>;, (const
struct X &) &_ZGR1x_;

    Since 'x' was declared constexpr, we do cxx_constant_init and we hit
    the preeval code added in r269003 while evaluating the INIT_EXPR:

      _ZGR1x_.x = (const struct X *) &<PLACEHOLDER_EXPR struct X> >>>

    but we have no ctx.ctor or ctx.object here so lookup_placeholder won't
    find anything that could replace X and we ICE on
     7861       /* A placeholder without a referent.  We can get here when
     7862          checking whether NSDMIs are noexcept, or in
massage_init_elt;
     7863          just say it's non-constant for now.  */
     7864       gcc_assert (ctx->quiet);
    because cxx_constant_init means !ctx->quiet.  It's not correct that
    there isn't a referent.  I think the following patch is a pretty
    straightforward fix: pass the _ZGR var down to maybe_constant_init so
    that it can replace the PLACEHOLDER_EXPR with _ZGR1x_.

    The commented assert in the test doesn't pass: we complain that _ZGR1x_
    isn't a constexpr variable because we don't implement DR2126 (PR101588).

            PR c++/107079

    gcc/cp/ChangeLog:

            * call.cc (set_up_extended_ref_temp): Pass var to
maybe_constant_init.

    gcc/testsuite/ChangeLog:

            * g++.dg/cpp0x/constexpr-nsdmi2.C: New test.

    (cherry picked from commit 67b82bc1b29b82e4577df9491793624f3a8eb12f)

[Bug c++/107079] [10/11 Regression] ICE initializing lifetime-extended constexpr variable that stores its this pointer

[mpolacek at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=107079

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

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|ASSIGNED                    |RESOLVED
         Resolution|---                         |FIXED
            Summary|[10/11/12 Regression] ICE   |[10/11 Regression] ICE
                   |initializing                |initializing
                   |lifetime-extended constexpr |lifetime-extended constexpr
                   |variable that stores its    |variable that stores its
                   |this pointer                |this pointer

--- Comment #9 from Marek Polacek <mpolacek at gcc dot gnu.org> ---
Fixed.

[Bug c++/107079] [10/11 Regression] ICE initializing lifetime-extended constexpr variable that stores its this pointer

[ppalka at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=107079

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

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |ppalka at gcc dot gnu.org
   Target Milestone|10.5                        |12.3