From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from eggs.gnu.org ([2001:4830:134:3::10]:37235) by lists.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1b1EC4-0007Ia-Fr for qemu-devel@nongnu.org; Fri, 13 May 2016 10:35:57 -0400 Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71) (envelope-from ) id 1b1EBy-00037V-Nv for qemu-devel@nongnu.org; Fri, 13 May 2016 10:35:55 -0400 Received: from mail-ig0-x233.google.com ([2607:f8b0:4001:c05::233]:36828) by eggs.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1b1EBy-000376-Cg for qemu-devel@nongnu.org; Fri, 13 May 2016 10:35:50 -0400 Received: by mail-ig0-x233.google.com with SMTP id qe5so9066242igc.1 for ; Fri, 13 May 2016 07:35:49 -0700 (PDT) References: <1462814989-24360-1-git-send-email-peter.maydell@linaro.org> <5732D667.3040105@huawei.com> <573493A7.3030604@linaro.org> <57349AC9.5070407@linaro.org> From: Shannon Zhao Message-ID: <5735E63B.7060500@linaro.org> Date: Fri, 13 May 2016 22:35:39 +0800 MIME-Version: 1.0 In-Reply-To: Content-Type: text/plain; charset=utf-8 Content-Transfer-Encoding: 8bit Subject: Re: [Qemu-devel] [PATCH 00/23] GICv3 emulation List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , To: Peter Maydell Cc: Shannon Zhao , qemu-arm , QEMU Developers , Patch Tracking , Shlomo Pongratz , Shlomo Pongratz , Pavel Fedin , Christoffer Dall On 2016年05月12日 23:22, Peter Maydell wrote: > On 12 May 2016 at 16:01, Shannon Zhao wrote: >> On 2016年05月12日 22:35, Peter Maydell wrote: >>> Can you put the Image file somewhere I can download it, please? >> I just upload it to >> http://people.linaro.org/~shannon.zhao/ACPI_ARM/Image > > OK, I can reproduce with that, and I think I know what this is. > If we boot a "-bios" image then we don't apply the QEMU workaround > for booting buggy Linux images which puts all the interrupts into > Group 1, and it looks like a no-TZ-support UEFI doesn't > configure the interrupt groups either. So Linux ends up hanging > because the interrupt is group 0 but group 0 is disabled. > > If you apply this patch to your guest kernel you should find > it boots OK I think: > https://lkml.org/lkml/2016/5/10/297 > Yeah, it boots OK with this patch. > (I'm not sure why the images I was running worked OK; I must > have been testing with an image that had that bugfix in it, > though I could have sworn I wasn't.) > > I don't really want to apply the QEMU workaround when we're > booting firmware images as well as raw kernels, but on the > other hand via-UEFI is kind of supposed to be the normal way > to boot in future. Perhaps we should drop the QEMU workaround > altogether, putting everybody on the same playing field... > Maybe and if people want to use old kernels, they could use the stable kernel after the kernel patch applied. Thanks, -- Shannon