From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1751690AbaIJNoG (ORCPT ); Wed, 10 Sep 2014 09:44:06 -0400 Received: from smtp.nue.novell.com ([195.135.221.5]:34029 "EHLO smtp.nue.novell.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751219AbaIJNoE (ORCPT ); Wed, 10 Sep 2014 09:44:04 -0400 Date: Wed, 10 Sep 2014 21:43:47 +0800 From: joeyli To: Yinghai Lu Cc: "Rafael J. Wysocki" , "Lee, Chun-Yi" , Thomas Gleixner , Ingo Molnar , "H. Peter Anvin" , Len Brown , Pavel Machek , Takashi Iwai , Linux PM list , Linux Kernel Mailing List Subject: Re: [PATCH] Hibernate: Do not assume the first e820 area to be RAM Message-ID: <20140910134347.GB18567@linux-rxt1.site> References: <1407754252-22779-1-git-send-email-jlee@suse.com> <1561567.ev2pEO8C5v@vostro.rjw.lan> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: 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 Hi Yinghai, Thanks for your review, first! On Tue, Sep 09, 2014 at 11:08:45PM -0700, Yinghai Lu wrote: > On Mon, Sep 8, 2014 at 3:52 PM, Rafael J. Wysocki wrote: > > On Monday, August 11, 2014 06:50:52 PM Lee, Chun-Yi wrote: > >> In arch/x86/kernel/setup.c::trim_bios_range(), the codes introduced > >> by 1b5576e6 (base on d8a9e6a5), it updates the first 4Kb of memory > >> to be E820_RESERVED region. That's because it's a BIOS owned area > >> but generally not listed in the E820 table: > >> > >> [ 0.000000] e820: BIOS-provided physical RAM map: > >> [ 0.000000] BIOS-e820: [mem 0x0000000000000000-0x0000000000096fff] usable > >> [ 0.000000] BIOS-e820: [mem 0x0000000000097000-0x0000000000097fff] reserved > >> ... > >> [ 0.000000] e820: update [mem 0x00000000-0x00000fff] usable ==> reserved > >> [ 0.000000] e820: remove [mem 0x000a0000-0x000fffff] usable > >> > >> But the region of first 4Kb didn't register to nosave memory: > >> > >> [ 0.000000] PM: Registered nosave memory: [mem 0x00097000-0x00097fff] > >> [ 0.000000] PM: Registered nosave memory: [mem 0x000a0000-0x000fffff] > >> > >> The codes in e820_mark_nosave_regions() assumes the first e820 area to be > >> RAM, so it causes the first 4Kb E820_RESERVED region ignored when register > >> to nosave. This patch removed assumption of the first e820 area. > >> > >> Cc: "Rafael J. Wysocki" > >> Cc: Len Brown > >> Cc: Pavel Machek > >> Cc: Thomas Gleixner > >> Cc: Ingo Molnar > >> Cc: "H. Peter Anvin" > >> Signed-off-by: Lee, Chun-Yi > > > > Thomas, Ingo, Peter, any objections here? > > > > If not, do you want to handle it or do you want me to do that? > > Did it address any regression? > I found this situation when comparing the e820 region with nosave memory address. But, I don't know any real machine which has bug report against this. > > > >> --- > >> arch/x86/kernel/e820.c | 7 +++---- > >> 1 file changed, 3 insertions(+), 4 deletions(-) > >> > >> diff --git a/arch/x86/kernel/e820.c b/arch/x86/kernel/e820.c > >> index 988c00a..d595240 100644 > >> --- a/arch/x86/kernel/e820.c > >> +++ b/arch/x86/kernel/e820.c > >> @@ -682,18 +682,17 @@ void __init parse_e820_ext(u64 phys_addr, u32 data_len) > >> * hibernation (32 bit) or software suspend and suspend to RAM (64 bit). > >> * > >> * This function requires the e820 map to be sorted and without any > >> - * overlapping entries and assumes the first e820 area to be RAM. > >> + * overlapping entries. > >> */ > >> void __init e820_mark_nosave_regions(unsigned long limit_pfn) > >> { > >> int i; > >> unsigned long pfn; > >> > >> - pfn = PFN_DOWN(e820.map[0].addr + e820.map[0].size); > >> - for (i = 1; i < e820.nr_map; i++) { > >> + for (i = 0; i < e820.nr_map; i++) { > >> struct e820entry *ei = &e820.map[i]; > >> > >> - if (pfn < PFN_UP(ei->addr)) > >> + if (i > 0 && pfn < PFN_UP(ei->addr)) > >> register_nosave_region(pfn, PFN_UP(ei->addr)); > > could avoid the i > 0 checking. > > >> > >> pfn = PFN_DOWN(ei->addr + ei->size); > >> > > > > following would be better ? > > @@ -682,15 +682,14 @@ void __init parse_e820_ext(u64 phys_addr, u32 data_len) > * hibernation (32 bit) or software suspend and suspend to RAM (64 bit). > * > * This function requires the e820 map to be sorted and without any > - * overlapping entries and assumes the first e820 area to be RAM. > + * overlapping entries. > */ > void __init e820_mark_nosave_regions(unsigned long limit_pfn) > { > int i; > - unsigned long pfn; > + unsigned long pfn = 0; > > - pfn = PFN_DOWN(e820.map[0].addr + e820.map[0].size); > - for (i = 1; i < e820.nr_map; i++) { > + for (i = 0; i < e820.nr_map; i++) { > struct e820entry *ei = &e820.map[i]; > > if (pfn < PFN_UP(ei->addr)) Yes, thanks for your suggestion, your change can avoid the i > 0 checking. I will send v2 patch to add your improvement. Thanks a lot! Joey Lee