public inbox for gcc-patches@gcc.gnu.org
 help / color / mirror / Atom feed
From: Jason Merrill <jason@redhat.com>
To: Marek Polacek <polacek@redhat.com>,
	GCC Patches <gcc-patches@gcc.gnu.org>
Subject: Re: [PATCH] c++: wrong error with static constexpr var in tmpl [PR109876]
Date: Fri, 26 May 2023 21:47:10 -0400	[thread overview]
Message-ID: <460bf6aa-a336-7ac2-0434-c8dcfd8d0b87@redhat.com> (raw)
In-Reply-To: <20230526231805.406578-1-polacek@redhat.com>

On 5/26/23 19:18, Marek Polacek wrote:
> Since r8-509, we'll no longer create a static temporary var for
> the initializer '{ 1, 2 }' for num in the attached test because
> the code in finish_compound_literal is now guarded by
> '&& fcl_context == fcl_c99' but it's fcl_functional here.  This
> causes us to reject num as non-constant when evaluating it in
> a template.
> 
> Jason's idea was to treat num as value-dependent even though it
> actually isn't.  This patch implements that suggestion.
> 
> The is_really_empty_class check is sort of non-obvious but the
> comment should explain why I added it.
> 
> Co-authored-by: Jason Merrill <jason@redhat.com>
> 
> Bootstrapped/regtested on x86_64-pc-linux-gnu, ok for trunk?
> 
> 	PR c++/109876
> 
> gcc/cp/ChangeLog:
> 
> 	* pt.cc (value_dependent_expression_p) <case VAR_DECL>: Treat a
> 	constexpr-declared non-constant variable as value-dependent.
> 
> gcc/testsuite/ChangeLog:
> 
> 	* g++.dg/cpp0x/constexpr-template12.C: New test.
> 	* g++.dg/cpp1z/constexpr-template1.C: New test.
> ---
>   gcc/cp/pt.cc                                  | 12 ++++++
>   .../g++.dg/cpp0x/constexpr-template12.C       | 38 +++++++++++++++++++
>   .../g++.dg/cpp1z/constexpr-template1.C        | 25 ++++++++++++
>   3 files changed, 75 insertions(+)
>   create mode 100644 gcc/testsuite/g++.dg/cpp0x/constexpr-template12.C
>   create mode 100644 gcc/testsuite/g++.dg/cpp1z/constexpr-template1.C
> 
> diff --git a/gcc/cp/pt.cc b/gcc/cp/pt.cc
> index 7fb3e75bceb..38fd8070705 100644
> --- a/gcc/cp/pt.cc
> +++ b/gcc/cp/pt.cc
> @@ -27969,6 +27969,18 @@ value_dependent_expression_p (tree expression)
>         else if (TYPE_REF_P (TREE_TYPE (expression)))
>   	/* FIXME cp_finish_decl doesn't fold reference initializers.  */
>   	return true;
> +      /* We have a constexpr variable and we're processing a template.  When
> +	 there's lifetime extension involved (for which finish_compound_literal
> +	 used to create a temporary), we'll not be able to evaluate the
> +	 variable until instantiating, so pretend it's value-dependent.  */
> +      else if (DECL_DECLARED_CONSTEXPR_P (expression)
> +	       && !TREE_CONSTANT (expression)
> +	       /* When there's nothing to initialize, we'll never mark the
> +		  VAR_DECL TREE_CONSTANT, therefore it would remain
> +		  value-dependent and we wouldn't instantiate.  */

Interesting.  Can we change that (i.e. mark it TREE_CONSTANT) rather 
than work around it here?

> +	       && !is_really_empty_class (TREE_TYPE (expression),
> +					  /*ignore_vptr*/false))
> +	return true;
>         return false;
>   
>       case DYNAMIC_CAST_EXPR:
> diff --git a/gcc/testsuite/g++.dg/cpp0x/constexpr-template12.C b/gcc/testsuite/g++.dg/cpp0x/constexpr-template12.C
> new file mode 100644
> index 00000000000..a9e065320c8
> --- /dev/null
> +++ b/gcc/testsuite/g++.dg/cpp0x/constexpr-template12.C
> @@ -0,0 +1,38 @@
> +// PR c++/109876
> +// { dg-do compile { target c++11 } }
> +
> +using size_t = decltype(sizeof 0);
> +
> +namespace std {
> +template <class> struct initializer_list {
> +  const int *_M_array;
> +  size_t _M_len;
> +  constexpr size_t size() const { return _M_len; }
> +};
> +} // namespace std
> +
> +constexpr std::initializer_list<int> gnum{2};
> +
> +template <int> struct Array {};
> +template <int> void g()
> +{
> +  static constexpr std::initializer_list<int> num{2};
> +  static_assert(num.size(), "");
> +  Array<num.size()> ctx;
> +
> +  constexpr Array<1> num1{};
> +}
> +
> +template <int N>
> +struct Foo
> +{
> +  static constexpr std::initializer_list<int> num = { 1, 2 };
> +  static_assert(num.size(), "");
> +  Array<num.size()> ctx;
> +};
> +
> +void
> +f (Foo<5>)
> +{
> +  g<0>();
> +}
> diff --git a/gcc/testsuite/g++.dg/cpp1z/constexpr-template1.C b/gcc/testsuite/g++.dg/cpp1z/constexpr-template1.C
> new file mode 100644
> index 00000000000..58be046fd36
> --- /dev/null
> +++ b/gcc/testsuite/g++.dg/cpp1z/constexpr-template1.C
> @@ -0,0 +1,25 @@
> +// PR c++/109876
> +// { dg-do compile { target c++17 } }
> +
> +struct Foo {};
> +template <const Foo&> struct X {};
> +
> +void g()
> +{
> +  static constexpr Foo foo;
> +  X<foo> x;
> +}
> +
> +template<int>
> +void f()
> +{
> +  static constexpr Foo foo;
> +  X<foo> x;
> +}
> +
> +void
> +h ()
> +{
> +  f<0>();
> +  f<1>();
> +}
> 
> base-commit: 8d6bd830f5f9c939e8565c0341a0c6c588834484


  reply	other threads:[~2023-05-27  1:47 UTC|newest]

Thread overview: 4+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2023-05-26 23:18 Marek Polacek
2023-05-27  1:47 ` Jason Merrill [this message]
2023-07-13 18:54   ` [PATCH v2] " Marek Polacek
2023-07-14 17:07     ` Jason Merrill

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=460bf6aa-a336-7ac2-0434-c8dcfd8d0b87@redhat.com \
    --to=jason@redhat.com \
    --cc=gcc-patches@gcc.gnu.org \
    --cc=polacek@redhat.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).