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 C3A2A3858CDA; Thu, 10 Nov 2022 16:56:36 +0000 (GMT)
DKIM-Filter: OpenDKIM Filter v2.11.0 sourceware.org C3A2A3858CDA
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gcc.gnu.org;
	s=default; t=1668099397;
	bh=CZq/KjCHTgNXmZWGHTsw+TYY5vbbvRpivcr943Z2mKs=;
	h=From:To:Subject:Date:In-Reply-To:References:From;
	b=SkZOq4ZAIs4oHTPpijYwXF5dAPmXvWNHWEXg4qEF5EeR9Kjux+nk/jNTlqftp5iVn
	 FUXGC1tH0Qk8AIwW/4LsiHKj2djSpDsImHfmnaQcZbs3nV31hN+QnExCMJgHTUWJCC
	 OIOYveoJmYP/oLchpOsbOTBS9MvajCcqkUfiJat0=
From: "pinskia at gcc dot gnu.org" <gcc-bugzilla@gcc.gnu.org>
To: gcc-bugs@gcc.gnu.org
Subject: [Bug target/107590] __atomic_test_and_set broken on PowerPC
Date: Thu, 10 Nov 2022 16:56:36 +0000
X-Bugzilla-Reason: CC
X-Bugzilla-Type: changed
X-Bugzilla-Watch-Reason: None
X-Bugzilla-Product: gcc
X-Bugzilla-Component: target
X-Bugzilla-Version: 11.3.0
X-Bugzilla-Keywords: 
X-Bugzilla-Severity: normal
X-Bugzilla-Who: pinskia at gcc dot gnu.org
X-Bugzilla-Status: WAITING
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-107590-4-qeRm2ll4Ym@http.gcc.gnu.org/bugzilla/>
In-Reply-To: <bug-107590-4@http.gcc.gnu.org/bugzilla/>
References: <bug-107590-4@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=3D107590

--- Comment #8 from Andrew Pinski <pinskia at gcc dot gnu.org> ---
(In reply to Peter Dimov from comment #7)=20
> I don't see an alignment requirement being mentioned here.

I think you misunderstood the alignment issue. There might be no alignment
requirement directly on __atomic_test_and_set but if there is a mismatch
understanding of alignment of what GCC thinks the alignment of the address
should be and what the address really is, then there will be an issue. now =
that
could be still a GCC bug or it can be a bug in the source; usually
suballocators providing wrong aligned addresses.


But there is not enough information here to figure out what exactly is going
wrong.
No preprocessed source of the code going wrong.
Not even register state or instructions where the segfault is happening (th=
is
is would be very useful but not required).=