From mboxrd@z Thu Jan 1 00:00:00 1970 From: Tony Lindgren Subject: Re: ARM errata 430973 on multi platform kernels Date: Fri, 3 Apr 2015 15:15:17 -0700 Message-ID: <20150403221517.GX10805@atomide.com> References: <1425236828-28349-1-git-send-email-jarkko.nikula@bitmer.com> <55197A12.1050009@bitmer.com> <20150330164237.GJ10805@atomide.com> <55198BA4.5010207@bitmer.com> <20150330175051.GK10805@atomide.com> <20150331123233.GA15103@earth> <20150401194734.GT10805@atomide.com> <20150403163553.GA16247@earth> <551F0F50.1030701@gmail.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Return-path: Received: from muru.com ([72.249.23.125]:41722 "EHLO muru.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752519AbbDCWTM (ORCPT ); Fri, 3 Apr 2015 18:19:12 -0400 Content-Disposition: inline In-Reply-To: <551F0F50.1030701@gmail.com> Sender: linux-omap-owner@vger.kernel.org List-Id: linux-omap@vger.kernel.org To: Ivaylo Dimitrov Cc: Sebastian Reichel , linux-arm-kernel@lists.infradead.org, linux-omap@vger.kernel.org, Pavel Machek * Ivaylo Dimitrov [150403 15:08]: > Hi > > On 3.04.2015 19:35, Sebastian Reichel wrote: > > > >Maybe an option would be to provide two cpu_v7_switch_mm > >implementations (one with the errata and one without). Then > >the system can start with the simple implementation. Once > >the boot as progressed far enough to know, that the hardware > >is affected by the errata, it could switch to the implementation > >with the flushing. > > Or rather the opposite, as if the kernel itself is thumb-compiled, it most > probably will have no chance to progress enough to switch to the correct > implementation before hanging. We should first verify the same bug happens with armel also. I just verified the CPU load in the background makes armhf apps segfault without $subject workaround enabled. If the segfaulting does not happen with armel, then chances it's some kind of neon related issue and the fix can be more targeted. Regards, Tony From mboxrd@z Thu Jan 1 00:00:00 1970 From: tony@atomide.com (Tony Lindgren) Date: Fri, 3 Apr 2015 15:15:17 -0700 Subject: ARM errata 430973 on multi platform kernels In-Reply-To: <551F0F50.1030701@gmail.com> References: <1425236828-28349-1-git-send-email-jarkko.nikula@bitmer.com> <55197A12.1050009@bitmer.com> <20150330164237.GJ10805@atomide.com> <55198BA4.5010207@bitmer.com> <20150330175051.GK10805@atomide.com> <20150331123233.GA15103@earth> <20150401194734.GT10805@atomide.com> <20150403163553.GA16247@earth> <551F0F50.1030701@gmail.com> Message-ID: <20150403221517.GX10805@atomide.com> To: linux-arm-kernel@lists.infradead.org List-Id: linux-arm-kernel.lists.infradead.org * Ivaylo Dimitrov [150403 15:08]: > Hi > > On 3.04.2015 19:35, Sebastian Reichel wrote: > > > >Maybe an option would be to provide two cpu_v7_switch_mm > >implementations (one with the errata and one without). Then > >the system can start with the simple implementation. Once > >the boot as progressed far enough to know, that the hardware > >is affected by the errata, it could switch to the implementation > >with the flushing. > > Or rather the opposite, as if the kernel itself is thumb-compiled, it most > probably will have no chance to progress enough to switch to the correct > implementation before hanging. We should first verify the same bug happens with armel also. I just verified the CPU load in the background makes armhf apps segfault without $subject workaround enabled. If the segfaulting does not happen with armel, then chances it's some kind of neon related issue and the fix can be more targeted. Regards, Tony