public inbox for gcc-bugs@sourceware.org help / color / mirror / Atom feed
From: "g.peterhoff@t-online.de" <gcc-bugzilla@gcc.gnu.org> To: gcc-bugs@gcc.gnu.org Subject: [Bug c++/109378] New: improve __builtin_sqrt Date: Sun, 02 Apr 2023 20:59:44 +0000 [thread overview] Message-ID: <bug-109378-4@http.gcc.gnu.org/bugzilla/> (raw) https://gcc.gnu.org/bugzilla/show_bug.cgi?id=109378 Bug ID: 109378 Summary: improve __builtin_sqrt Product: gcc Version: unknown Status: UNCONFIRMED Severity: normal Priority: P3 Component: c++ Assignee: unassigned at gcc dot gnu.org Reporter: g.peterhoff@t-online.de Target Milestone: --- Hello gcc team, https://godbolt.org/z/Wa1rfxrPo when I write a function that contains std::sqrt, it always contains the nan(?) tests for the argument. E.g. sqrtf64. If I use my_sqrt the tests are done inside sqrt and not in the calling function - clear (because noinline). Wouldn't it be better to rewrite __builtin_sqrt so that these tests are done inside __builtin_sqrt and not already in the calling context? This would have the advantage that std::sqrt would not "contaminate" the calling function with conditional jumps and thus inflate it. I can make this clear with foo vs. bar. And of course __builtin_sqrt must be able to be vectorized automatically and must be inline for certain contexts (e.g. __FAST_MATH__). regards Gero
next reply other threads:[~2023-04-02 20:59 UTC|newest] Thread overview: 14+ messages / expand[flat|nested] mbox.gz Atom feed top 2023-04-02 20:59 g.peterhoff@t-online.de [this message] 2023-04-02 21:01 ` [Bug c++/109378] " pinskia at gcc dot gnu.org 2023-04-02 21:23 ` g.peterhoff@t-online.de 2023-04-02 21:27 ` pinskia at gcc dot gnu.org 2023-04-02 21:51 ` g.peterhoff@t-online.de 2023-04-02 21:53 ` pinskia at gcc dot gnu.org 2023-04-02 21:54 ` [Bug c++/109378] new builtin like __builtin_sqrt but does not set errno pinskia at gcc dot gnu.org 2023-04-02 21:57 ` pinskia at gcc dot gnu.org 2023-04-02 22:01 ` pinskia at gcc dot gnu.org 2023-04-02 22:39 ` g.peterhoff@t-online.de 2023-04-02 22:55 ` pinskia at gcc dot gnu.org 2023-04-02 22:57 ` pinskia at gcc dot gnu.org 2023-04-03 0:01 ` g.peterhoff@t-online.de 2023-04-03 0:39 ` pinskia 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-109378-4@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).