From mboxrd@z Thu Jan  1 00:00:00 1970
Return-Path: <gcc-bugzilla@gcc.gnu.org>
Received: by sourceware.org (Postfix, from userid 48)
	id 24D7538582A3; Wed,  6 Mar 2024 12:35:43 +0000 (GMT)
DKIM-Filter: OpenDKIM Filter v2.11.0 sourceware.org 24D7538582A3
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gcc.gnu.org;
	s=default; t=1709728543;
	bh=+/8y4PeT1CeU0byjgB1O9iPwsGoUVwL0bMKefqj5wUc=;
	h=From:To:Subject:Date:In-Reply-To:References:From;
	b=wa5GEMF36s+IGrwl96QwmKYq+NyNtz7yBXF4UFbSA0kYwxizhqazHUIvcr/GIBSUM
	 NECA6JO7xsAwt2XwjmyZHIeyXThZ9U0d2QqchofVBjoCVPQ08CAObKEfWphfI0wbcN
	 1vyoz9sIVGgFgU0+7+vNsxd4ZwAYeRc+LdME8B7M=
From: "redi at gcc dot gnu.org" <gcc-bugzilla@gcc.gnu.org>
To: gcc-bugs@gcc.gnu.org
Subject: [Bug libstdc++/77776] C++17 std::hypot implementation is poor
Date: Wed, 06 Mar 2024 12:35:42 +0000
X-Bugzilla-Reason: CC
X-Bugzilla-Type: changed
X-Bugzilla-Watch-Reason: None
X-Bugzilla-Product: gcc
X-Bugzilla-Component: libstdc++
X-Bugzilla-Version: 7.0
X-Bugzilla-Keywords: 
X-Bugzilla-Severity: normal
X-Bugzilla-Who: redi at gcc dot gnu.org
X-Bugzilla-Status: ASSIGNED
X-Bugzilla-Resolution: 
X-Bugzilla-Priority: P3
X-Bugzilla-Assigned-To: emsr at gcc dot gnu.org
X-Bugzilla-Target-Milestone: ---
X-Bugzilla-Flags: 
X-Bugzilla-Changed-Fields: 
Message-ID: <bug-77776-4-4UD9Ch5pYO@http.gcc.gnu.org/bugzilla/>
In-Reply-To: <bug-77776-4@http.gcc.gnu.org/bugzilla/>
References: <bug-77776-4@http.gcc.gnu.org/bugzilla/>
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable
X-Bugzilla-URL: http://gcc.gnu.org/bugzilla/
Auto-Submitted: auto-generated
MIME-Version: 1.0
List-Id: <gcc-bugs.sourceware.org>

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=3D77776

--- Comment #21 from Jonathan Wakely <redi at gcc dot gnu.org> ---
(In reply to g.peterhoff from comment #19)
> * You were probably wondering why I wrote "if (std::isinf(x) | std::isinf=
(y)
> | std::isinf(z))", for example. This is intentional. The problem is that =
gcc
> almost always produces branch code for logical operations, so *a lot* of
> conditional jumps. By using arithmetic operations, so instead of || && ju=
st
> | &, I can get it to generate only actually necessary conditional jumps or
> cmoves. branchfree code is always better.

In general the code is not equivalent if any of the subexpressions has side
effects, such as setting errno, raising an exception for a signalling nan e=
tc.
but I don't think we care about that here.=