From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S269760AbUJALaF (ORCPT ); Fri, 1 Oct 2004 07:30:05 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S269761AbUJALaF (ORCPT ); Fri, 1 Oct 2004 07:30:05 -0400 Received: from gprs214-247.eurotel.cz ([160.218.214.247]:640 "EHLO amd.ucw.cz") by vger.kernel.org with ESMTP id S269760AbUJAL3A (ORCPT ); Fri, 1 Oct 2004 07:29:00 -0400 Date: Fri, 1 Oct 2004 13:28:47 +0200 From: Pavel Machek To: torvalds@transmeta.com, Andrew Morton , Patrick Mochel , kernel list Subject: swsusp: fix highmem corruption in -rc3 Message-ID: <20041001112846.GA1489@elf.ucw.cz> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline X-Warning: Reading this can be dangerous to your mental health. User-Agent: Mutt/1.5.5.1+cvs20040105i Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org Hi! -rc3 swsusp breaks badly with highmem, because patch from -mm is missing. Please apply this patch from -mm to vanilla. Otherwise -rc3 swsusp seems to work here... Quite good result for so big merge. Thanks. (But this certainly scared me a lot). Pavel --- clean/kernel/power/swsusp.c 2004-10-01 00:30:32.000000000 +0200 +++ clean-mm/kernel/power/swsusp.c 2004-09-26 01:34:27.000000000 +0200 @@ -856,7 +860,9 @@ local_irq_disable(); save_processor_state(); error = swsusp_arch_suspend(); + /* Restore control flow magically appears here */ restore_processor_state(); + restore_highmem(); local_irq_enable(); return error; } -- People were complaining that M$ turns users into beta-testers... ...jr ghea gurz vagb qrirybcref, naq gurl frrz gb yvxr vg gung jnl!