public inbox for gcc-bugs@sourceware.org help / color / mirror / Atom feed
From: "redi at gcc dot gnu.org" <gcc-bugzilla@gcc.gnu.org> To: gcc-bugs@gcc.gnu.org Subject: [Bug c++/104091] -std=c++20 causing meaningless error message "'auto' not allowed in alias declaration" which should be "missing template arguments after ..." Date: Mon, 07 Nov 2022 09:49:28 +0000 [thread overview] Message-ID: <bug-104091-4-zT6bBt1lvj@http.gcc.gnu.org/bugzilla/> (raw) In-Reply-To: <bug-104091-4@http.gcc.gnu.org/bugzilla/> https://gcc.gnu.org/bugzilla/show_bug.cgi?id=104091 Jonathan Wakely <redi at gcc dot gnu.org> changed: What |Removed |Added ---------------------------------------------------------------------------- Last reconfirmed|2022-01-18 00:00:00 |2022-11-7 --- Comment #2 from Jonathan Wakely <redi at gcc dot gnu.org> --- I've just hit this myself and was quite confused by the diagnostic. A realistic but still close to minimal reproducer is: namespace x { template<class T> struct S { }; } struct A { using S = ::x::S; }; Unsurprisingly, you get the same error with -std=c++17 -fconcepts, because the compiler seems to be treating the template-id as a placeholder-type-specifier, maybe even as a type-constraint. Neither makes sense as the defining-type-id of an alias-declaration. With -std=c++14 you get a correct "invalid use of template-name" error, but then an incorrect note about CTAD: using.cc:8:13: error: invalid use of template-name 'x::S' without an argument list 8 | using S = ::x::S; | ^~ using.cc:8:13: note: class template argument deduction is only available with '-std=c++17' or '-std=gnu++17' using.cc:3:28: note: 'template<class T> struct x::S' declared here 3 | template<class T> struct S { }; | ^ Although it's true that CTAD isn't available in C++14, it's not relevant here because it couldn't be used here anyway (as evidenced by the fact it still doesn't compie in C++17). I would expect the same error for C++14 and C++17 (and ideally C++20 too) since the code is invalid in exactly the same way in all cases. C++17 gives a similar error to C++14 but with slightly different wording for some reason: using.cc:8:13: error: missing template arguments after 'x::S' 8 | using S = ::x::S; | ^~ | <> using.cc:3:28: note: 'template<class T> struct x::S' declared here 3 | template<class T> struct S { }; | ^ So we should: 1) fix the bogus "auto not allowed" when concepts are enabled; 2) harmonize the C++14 and C++17 wording about missing template args; 3) remove the bogus note about CTAD.
next prev parent reply other threads:[~2022-11-07 9:49 UTC|newest] Thread overview: 5+ messages / expand[flat|nested] mbox.gz Atom feed top 2022-01-18 11:00 [Bug c++/104091] New: " nickhuang99 at hotmail dot com 2022-01-18 11:10 ` [Bug c++/104091] " pinskia at gcc dot gnu.org 2022-11-07 9:49 ` redi at gcc dot gnu.org [this message] 2022-11-07 9:53 ` redi at gcc dot gnu.org 2022-11-07 9:55 ` redi at gcc dot gnu.org
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=bug-104091-4-zT6bBt1lvj@http.gcc.gnu.org/bugzilla/ \ --to=gcc-bugzilla@gcc.gnu.org \ --cc=gcc-bugs@gcc.gnu.org \ /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: linkBe 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).