From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1755138Ab2HOOuP (ORCPT ); Wed, 15 Aug 2012 10:50:15 -0400 Received: from youngberry.canonical.com ([91.189.89.112]:35144 "EHLO youngberry.canonical.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1753366Ab2HOOuM (ORCPT ); Wed, 15 Aug 2012 10:50:12 -0400 Date: Wed, 15 Aug 2012 11:50:01 -0300 From: Herton Ronaldo Krzesinski To: Greg Kroah-Hartman Cc: linux-kernel@vger.kernel.org, stable@vger.kernel.org, torvalds@linux-foundation.org, akpm@linux-foundation.org, alan@lxorguk.ukuu.org.uk, Colin Cross , Barry Song , Catalin Marinas , Ido Yariv , Daniel Drake , Will Deacon , Russell King Subject: Re: [ 07/44] ARM: 7477/1: vfp: Always save VFP state in vfp_pm_suspend on UP Message-ID: <20120815145000.GA3323@herton-Z68MA-D2H-B3> References: <20120813220142.113186818@linuxfoundation.org> <20120813220142.801309366@linuxfoundation.org> <20120814200110.GE5277@herton-Z68MA-D2H-B3> <20120815140541.GB22871@kroah.com> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <20120815140541.GB22871@kroah.com> User-Agent: Mutt/1.5.21 (2010-09-15) Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org On Wed, Aug 15, 2012 at 07:05:41AM -0700, Greg Kroah-Hartman wrote: > On Tue, Aug 14, 2012 at 05:01:11PM -0300, Herton Ronaldo Krzesinski wrote: > > On Mon, Aug 13, 2012 at 03:02:14PM -0700, Greg Kroah-Hartman wrote: > > > From: Greg KH > > > > > [...] > > > --- a/arch/arm/vfp/vfpmodule.c > > > +++ b/arch/arm/vfp/vfpmodule.c > > > @@ -412,6 +412,12 @@ static int vfp_pm_suspend(void) > > > > > > /* disable, just in case */ > > > fmxr(FPEXC, fmrx(FPEXC) & ~FPEXC_EN); > > > + } else if (vfp_current_hw_state[ti->cpu]) { > > > +#ifndef CONFIG_SMP > > > + fmxr(FPEXC, fpexc | FPEXC_EN); > > > + vfp_save_state(vfp_current_hw_state[ti->cpu], fpexc); > > > + fmxr(FPEXC, fpexc); > > > +#endif > > > > This fails to build on 3.0: > > > > linux-stable/arch/arm/vfp/vfpmodule.c: In function 'vfp_pm_suspend': > > linux-stable/arch/arm/vfp/vfpmodule.c:415:13: error: 'vfp_current_hw_state' undeclared (first use in this function) > > linux-stable/arch/arm/vfp/vfpmodule.c:415:13: note: each undeclared identifier is reported only once for each function it appears in > > make[3]: *** [arch/arm/vfp/vfpmodule.o] Error 1 > > > > In 3.0, vfp_current_hw_state was still named last_VFP_context, so I > > think just renaming should be enough to fix this. > > I'll remove the patch as it probably just shouldn't be there, right? If > you think it should be in 3.0-stable, can you please send me a fixed up > patch that you have tested that it at least builds properly? I'm not intimate with this ARM code, so I think someone else in Cc can answer better, I just saw it doesn't build with 3.0. If the same problem affects 3.0, then the change should be backported renaming vfp_current_hw_state to last_VFP_context. It seems the same code was there with vfp_current_hw_state still named as last_VFP_context, so looks like 3.0 would be affected as well. Also I noted that because the same naming, 3.0 didn't got commit a84b895a2348f0dbff31b71ddf954f70a6cde368 ("ARM: 7476/1: vfp: only clear vfp state for current cpu in vfp_pm_suspend") as the patch would fail to apply on 3.0 code base, while 3.4.9 and 3.5.2 have it. I assume someone who did these commits or know it/are affected can take a look and if needed backport them to 3.0? > > thanks, > > greg k-h > -- []'s Herton