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 47C163858C66; Tue,  6 Feb 2024 22:14:45 +0000 (GMT)
DKIM-Filter: OpenDKIM Filter v2.11.0 sourceware.org 47C163858C66
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gcc.gnu.org;
	s=default; t=1707257685;
	bh=q+KJrBiEc2ULc4gun9SjI026F1VSI1SBQcZzUwwGwkg=;
	h=From:To:Subject:Date:In-Reply-To:References:From;
	b=YjcbXX59ApkXfgvj8nfC7xDQWtDaaQE3ioky+KgF1gzWnUiNUWTV1QzUe06ty6fLm
	 97j2rVtAbrovZhiCEOHJpzOV5LaMkeOD/6qNMI5sXV2Fj+YRNDkYFm+uCoRMDWUgGV
	 ubu0j5+UZ1wIIuVoj3f7Hw+oaDp1ksAe3yTee5d8=
From: "jakub at gcc dot gnu.org" <gcc-bugzilla@gcc.gnu.org>
To: gcc-bugs@gcc.gnu.org
Subject: [Bug libstdc++/87744] Some valid instantiations of
 linear_congruential_engine produce compiler errors when __int128 isn't
 available
Date: Tue, 06 Feb 2024 22:14:44 +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: 7.3.0
X-Bugzilla-Keywords: rejects-valid
X-Bugzilla-Severity: normal
X-Bugzilla-Who: jakub at gcc dot gnu.org
X-Bugzilla-Status: NEW
X-Bugzilla-Resolution: 
X-Bugzilla-Priority: P3
X-Bugzilla-Assigned-To: unassigned at gcc dot gnu.org
X-Bugzilla-Target-Milestone: ---
X-Bugzilla-Flags: 
X-Bugzilla-Changed-Fields: 
Message-ID: <bug-87744-4-r0aZCtdWOD@http.gcc.gnu.org/bugzilla/>
In-Reply-To: <bug-87744-4@http.gcc.gnu.org/bugzilla/>
References: <bug-87744-4@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=3D87744

--- Comment #11 from Jakub Jelinek <jakub at gcc dot gnu.org> ---
(In reply to Jakub Jelinek from comment #10)
> As discussed on IRC, probably the
> if (!__builtin_mul_overflow(__l._M_lo, __x, &__lo))
> optimization isn't a good idea, because most likely the compiler will in
> that case
> expand that to a full 64x64->128 multiplication plus testing that the upp=
er
> 64 bits are or aren't zero, so computing everything this function needs a=
nd
> doing it again if there is overflow, without possibility to query what it
> computed.

Though, if it is common enough, one could try to optimize the __ll[0] =3D=
=3D 0 &&
__xx[0] =3D=3D 0 case, one can do then either 32x32->64 or 64x64->64 multip=
lication
and be done with it.  But if it is rare in random's usage, it would just ma=
ke
the code larger.=