From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: by sourceware.org (Postfix, from userid 48) id AE70C38582A3; Tue, 7 Mar 2023 16:27:57 +0000 (GMT) DKIM-Filter: OpenDKIM Filter v2.11.0 sourceware.org AE70C38582A3 DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gcc.gnu.org; s=default; t=1678206477; bh=+3zPcvXsGZcG2UsgDXtQ58CrPSaNlhev8OzTmgktrpY=; h=From:To:Subject:Date:In-Reply-To:References:From; b=La5CRm8VetOzRrbLwFbMlgA0bOfybluHiKMMLyawOp6AEPDr+2cN9thXppH5NOyOk 83GlsBqy6gaJRKi13c3ba6WgziQpy4umlX3h59iQYNW/qPk2sA0rm/sPQTPRsyF4qq 5g5N4j/ZZpFK70qv+p3efhBBHjebMpLv2JxC62wA= From: "iains at gcc dot gnu.org" To: gcc-bugs@gcc.gnu.org Subject: [Bug c++/101118] coroutines: unexpected ODR warning for coroutine frame type in LTO builds Date: Tue, 07 Mar 2023 16:27:57 +0000 X-Bugzilla-Reason: CC X-Bugzilla-Type: changed X-Bugzilla-Watch-Reason: None X-Bugzilla-Product: gcc X-Bugzilla-Component: c++ X-Bugzilla-Version: 11.1.0 X-Bugzilla-Keywords: C++-coroutines, diagnostic, lto X-Bugzilla-Severity: normal X-Bugzilla-Who: iains 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: In-Reply-To: References: 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: https://gcc.gnu.org/bugzilla/show_bug.cgi?id=3D101118 --- Comment #16 from Iain Sandoe --- (In reply to Jan Hubicka from comment #15) > > https://gcc.gnu.org/bugzilla/show_bug.cgi?id=3D101118 > >=20 > > --- Comment #14 from Iain Sandoe --- > > (In reply to Jan Hubicka from comment #13) > > > > So .. for promotion of target expression temporaries to frame vars,= one of: > > > > - a) we need to find a different way to name them > > > I think we can just count number of fields within a given frame type? > >=20 > > yeah I made a hack that did this (and resolves this PR) but I'd think w= e can > > find something neater, I'd like to c++-ify the sources some more, and c= reate a > > class to manage the frame... ( maybe for GCC14 now ). > Thanks! > Since this is really part of the ABI, I wonder why it is not covered by > the IA-64 C++ ABI? Was it ever updated for coroutines? There is a coroutines ABI agreed between the "vendors" I have editorship of= the draft - and I suppose it needs to be included in the Itanium (and MSVC) ABI docs. However, the only part of the ABI that is described by this is the type-era= sed interface (there is never a need for the body of the coroutine, and detailed frame layout) to be interpreted by a coroutine body from a second compiler.= ** other than the portion of the frame that contains the pointers to resume and destroy **. The actual code is TU-local (and must remain that way even under LTO)=