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 582AA385AC2D; Tue, 25 Oct 2022 16:08:44 +0000 (GMT)
DKIM-Filter: OpenDKIM Filter v2.11.0 sourceware.org 582AA385AC2D
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gcc.gnu.org;
	s=default; t=1666714124;
	bh=9A2kJ65a/ibjjaLwX61ijAzH21/eRJvysPh4Jtwonqo=;
	h=From:To:Subject:Date:In-Reply-To:References:From;
	b=gOUOTwXaMCxAxVidIOin4J+2FTzMAVRNCm9AmnRiKMNpO5dRXSdqqUFLRk9cEpvuP
	 ZN7HTDHfNiT0Z/GlmuQzIOTLnfwzVrXoyP2VXM34JP2gEmWM8eRnrZyc9vBMDYgSe/
	 aTJ8roX/+ReMDUVvQAx0Lwv9dIJvOyQK0Dvvu7gQ=
From: "amonakov at gcc dot gnu.org" <gcc-bugzilla@gcc.gnu.org>
To: gcc-bugs@gcc.gnu.org
Subject: [Bug other/107353] [13 regression] Numerous ICEs after
 r13-3416-g1d561e1851c466
Date: Tue, 25 Oct 2022 16:08:44 +0000
X-Bugzilla-Reason: CC
X-Bugzilla-Type: changed
X-Bugzilla-Watch-Reason: None
X-Bugzilla-Product: gcc
X-Bugzilla-Component: other
X-Bugzilla-Version: 13.0
X-Bugzilla-Keywords: 
X-Bugzilla-Severity: normal
X-Bugzilla-Who: amonakov at gcc dot gnu.org
X-Bugzilla-Status: NEW
X-Bugzilla-Resolution: 
X-Bugzilla-Priority: P1
X-Bugzilla-Assigned-To: unassigned at gcc dot gnu.org
X-Bugzilla-Target-Milestone: 13.0
X-Bugzilla-Flags: 
X-Bugzilla-Changed-Fields: 
Message-ID: <bug-107353-4-ecRZOql1HY@http.gcc.gnu.org/bugzilla/>
In-Reply-To: <bug-107353-4@http.gcc.gnu.org/bugzilla/>
References: <bug-107353-4@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=3D107353

--- Comment #12 from Alexander Monakov <amonakov at gcc dot gnu.org> ---
ICE on the emutls-3.c testcase isn't related to emutls. Rather, the frontend
invokes decl_default_tls_model before attributes are processed, so the first
time around we miss the 'common' attribute when deciding the TLS access mod=
el.

The following cut-down testcase fails on x86 as well with -m32 -fpie:

__attribute__((common))
__thread int i;

int *f()
{
        return &i;
}

Before the offending commit GCC compiled 'f' as if the attribute was ignore=
d.
(on ELF targets combining TLS and COMMON is problematic if not undefined)=