From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from eggs.gnu.org ([2001:4830:134:3::10]:46494) by lists.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1bK9w5-0008Pr-N5 for qemu-devel@nongnu.org; Mon, 04 Jul 2016 15:53:42 -0400 Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71) (envelope-from ) id 1bK9w1-0001GW-4O for qemu-devel@nongnu.org; Mon, 04 Jul 2016 15:53:40 -0400 Received: from mx1.redhat.com ([209.132.183.28]:48323) by eggs.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1bK9w0-0001GM-VA for qemu-devel@nongnu.org; Mon, 04 Jul 2016 15:53:37 -0400 Received: from int-mx11.intmail.prod.int.phx2.redhat.com (int-mx11.intmail.prod.int.phx2.redhat.com [10.5.11.24]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by mx1.redhat.com (Postfix) with ESMTPS id 001AA461D2 for ; Mon, 4 Jul 2016 19:53:35 +0000 (UTC) Message-ID: <1467662012.15123.187.camel@redhat.com> From: Gerd Hoffmann Date: Mon, 04 Jul 2016 21:53:32 +0200 In-Reply-To: <1539298406.2095976.1467652793714.JavaMail.zimbra@redhat.com> References: <20160704153823.16879-1-marcandre.lureau@redhat.com> <20160704163139.GM3763@redhat.com> <20160704170715.GO3763@redhat.com> <0e0b3d90-24c2-519f-7b08-2b1174b35dc9@redhat.com> <1539298406.2095976.1467652793714.JavaMail.zimbra@redhat.com> Content-Type: text/plain; charset="UTF-8" Content-Transfer-Encoding: quoted-printable Mime-Version: 1.0 Subject: Re: [Qemu-devel] [PATCH] char: do not use atexit cleanup handler List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , To: =?ISO-8859-1?Q?Marc-Andr=E9?= Lureau Cc: Paolo Bonzini , "Daniel P. Berrange" , marcandre lureau , qemu-devel@nongnu.org, Frediano Ziglio Hi, > > > What about graphics threads ? In particular I'd be thinking of spice > > > which uses threads and chardevs. > >=20 > > I think it should be quiesced after pause_all_vcpus returns. Marc-Andr= =C3=A9 > > should know, but it's better to check with Gerd. >=20 > In theory, spice_server_vm_stop() should be called at this point, Yes, that should handle the qxl worker thread. > and all chardev in spice are stopped too there, as well as the qxl > worker processing thread (although the thread is not joined here > neither..). The chardevs are handled in iothread context anyway, so I don't think they need any special care. cheers, Gerd