From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1755353AbbA2V35 (ORCPT ); Thu, 29 Jan 2015 16:29:57 -0500 Received: from smtp.codeaurora.org ([198.145.11.231]:53111 "EHLO smtp.codeaurora.org" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1750944AbbA2V3z (ORCPT ); Thu, 29 Jan 2015 16:29:55 -0500 Message-ID: <54CAA651.2010205@codeaurora.org> Date: Thu, 29 Jan 2015 13:29:53 -0800 From: Stephen Boyd User-Agent: Mozilla/5.0 (X11; Linux i686 on x86_64; rv:31.0) Gecko/20100101 Thunderbird/31.4.0 MIME-Version: 1.0 To: Catalin Marinas CC: Russell King , "linux-arm-msm@vger.kernel.org" , "linux-kernel@vger.kernel.org" , "linux-arm-kernel@lists.infradead.org" , Kumar Gala Subject: Re: [PATCH 0/2] Fix secondary CPU boot for thumb2 kernels References: <1421869698-27341-1-git-send-email-sboyd@codeaurora.org> <54C94511.30508@codeaurora.org> <20150129154624.GH8951@e104818-lin.cambridge.arm.com> In-Reply-To: <20150129154624.GH8951@e104818-lin.cambridge.arm.com> Content-Type: text/plain; charset=windows-1252 Content-Transfer-Encoding: 7bit Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org On 01/29/15 07:46, Catalin Marinas wrote: > On Wed, Jan 28, 2015 at 08:22:41PM +0000, Stephen Boyd wrote: >> On 01/21/15 11:48, Stephen Boyd wrote: >>> Currently I can't boot up secondary CPUs on qcom platforms when I >>> compile the kernel for THUMB2 mode. This is because we always enter >>> the kernel in ARM mode regardless of what mode the kernel is compiled >>> for. This patchset adds a small wrapper to secondary_startup() called >>> secondary_startup_arm() that allows us to switch into THUMB2 mode if >>> we need to. If the kernel is compiled for ARM mode it shouldn't have >>> any effect. >>> >>> Stephen Boyd (2): >>> ARM: Add a secondary_startup that assumes ARM mode >>> ARM: qcom: Use secondary_startup_arm() >>> >>> arch/arm/kernel/head.S | 7 +++++++ >>> arch/arm/mach-qcom/platsmp.c | 4 ++-- >>> 2 files changed, 9 insertions(+), 2 deletions(-) >> Any comments here? Perhaps I can send this through the patch tracker? > FWIW, they look fine to me. > > Acked-by: Catalin Marinas > > I guess the reason it works fine on other platforms is because the > secondary start address we pass to firmware has bit 0 set and this > causes the jump to the kernel in Thumb-2 mode. > Thanks. I tried forcing that and it still didn't work so I guess the firmware isn't using interworking branches. -- Qualcomm Innovation Center, Inc. is a member of Code Aurora Forum, a Linux Foundation Collaborative Project