From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from eggs.gnu.org ([2001:4830:134:3::10]:44816) by lists.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1gPYoM-00057i-Kl for qemu-devel@nongnu.org; Wed, 21 Nov 2018 15:09:27 -0500 Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71) (envelope-from ) id 1gPYfM-00053Y-1Z for qemu-devel@nongnu.org; Wed, 21 Nov 2018 15:00:07 -0500 Received: from mail1.windriver.com ([147.11.146.13]:51565) by eggs.gnu.org with esmtps (TLS1.0:DHE_RSA_AES_256_CBC_SHA1:32) (Exim 4.71) (envelope-from ) id 1gPYfL-0004wO-PX for qemu-devel@nongnu.org; Wed, 21 Nov 2018 15:00:03 -0500 Received: from ALA-HCA.corp.ad.wrs.com ([147.11.189.40]) by mail1.windriver.com (8.15.2/8.15.1) with ESMTPS id wALHcJYU004539 (version=TLSv1 cipher=AES128-SHA bits=128 verify=FAIL) for ; Wed, 21 Nov 2018 09:38:19 -0800 (PST) References: <87pnvn6hnq.fsf@linaro.org> <87lg6b5shx.fsf@linaro.org> <98428a27-8634-9521-297c-a61c618e867a@windriver.com> <20181102175111.GA2388@work-vm> <54304edc-9de4-a25a-63d0-96518c5d7956@windriver.com> From: Chris Friesen Message-ID: Date: Wed, 21 Nov 2018 11:38:16 -0600 MIME-Version: 1.0 In-Reply-To: <54304edc-9de4-a25a-63d0-96518c5d7956@windriver.com> Content-Type: text/plain; charset="utf-8"; format=flowed Content-Language: en-US Content-Transfer-Encoding: quoted-printable Subject: Re: [Qemu-devel] strange situation, guest cpu thread spinning at ~100%, but display not yet initialized List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , To: qemu-devel@nongnu.org On 11/2/2018 2:45 PM, Chris Friesen wrote: > On 11/2/2018 11:51 AM, Dr. David Alan Gilbert wrote: >> =C2=A0=C2=A0=C2=A0 so the fix is Fam's 'aio: Do aio_notify_accept only= during blocking >> aio_poll'.=C2=A0 I see you're running the qemu-kvm-ev from centos, if = I read >> the version tea-leaves right, then I think that patch is in the >> 2.10.0-21.el7_5.7.1 package I can see. >=20 > It looks like we do not have this fix in our code, will definitely be=20 > giving it a try. Well, it appears that this was not the fix. With the patches applied we=20 still saw the problem. One of my coworkers has been investigating further, and he sent me the=20 following information: "One of the things the BIOS does before initializing the VGA is to start=20 the CPUs. More specifically, from CPU-0, it broadcasts a Setup=20 InterProcess Interrupt (SIPI) to all the remaining processors so that=20 they wake up and are initialized. CPU-0 busy-waits for all the=20 Instance=E2=80=99s CPUs to be initialized, and once that happens, it move= s on to=20 continue booting (which includes initializing the VGA). In our case,=20 one CPU fails to respond to the SIPI causing CPU-0 to keep busy waiting=20 for it. I am in the process of adding more instrumentation to qemu so=20 that we can know whether the missing CPU did not receive the SIPI or=20 what happened." Chris