Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1753144Ab1DAXFn (ORCPT ); Fri, 1 Apr 2011 19:05:43 -0400 Received: from rcsinet10.oracle.com ([148.87.113.121]:23089 "EHLO rcsinet10.oracle.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751713Ab1DAXFm (ORCPT ); Fri, 1 Apr 2011 19:05:42 -0400 Message-ID: <4D965A08.2080203@kernel.org> Date: Fri, 01 Apr 2011 16:04:40 -0700 From: Yinghai Lu User-Agent: Mozilla/5.0 (X11; U; Linux x86_64; en-US; rv:1.9.2.14) Gecko/20110221 SUSE/3.1.8 Thunderbird/3.1.8 MIME-Version: 1.0 To: "H. Peter Anvin" CC: Stefano Stabellini , Ingo Molnar , "Rafael J. Wysocki" , Michael Leun , "linux-kernel@vger.kernel.org" , Greg Kroah-Hartman , Mike Pagano Subject: Re: 2.6.38.2 breaks suspend to disk References: <20110330203215.0a1a41a7@xenia.leun.net> <20110331090524.07ad0069@xenia.leun.net> <201103312348.53678.rjw@sisk.pl> <4D94FE37.8070109@kernel.org> <4D95F80D.7070201@kernel.org> <4D95FBDD.9050901@zytor.com> <4D961FBC.2030105@zytor.com> <4D962837.2070300@kernel.org> <4D96294B.5050909@zytor.com> <4D962D72.2010501@kernel.org> <4D9633E7.1010707@zytor.com> <4D9643F1.1090807@zytor.com> <4D9646C6.7060301@zytor.com> In-Reply-To: <4D9646C6.7060301@zytor.com> Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit X-Source-IP: acsmt358.oracle.com [141.146.40.158] X-Auth-Type: Internal IP X-CT-RefId: str=0001.0A090209.4D965A1A.0041,ss=1,fgs=0 Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Length: 4112 Lines: 128 On 04/01/2011 02:42 PM, H. Peter Anvin wrote: > > And why on Earth is it worth saving a couple of instructions (and > introducing code ugliness and a more complex testing matrix) in the case > when it is not? Please check this one, it moves storing mmu_cr4 to arch_prepare_suspend. Thanks Yinghai From: Stefano Stabellini [PATCH -v5] x86: Save cr4 to mmu_cr4_features at boot time Save cr4 to mmu_cr4_features at boot time Michael reported 2.6.38.2 hibernation is broken by one backported patch. it cause a freeze when resuming from hibernation | "x86: Cleanup highmap after brk is concluded" | commit id e5f15b45ddf3afa2bbbb10c7ea34fb32b6de0a0e. it turns out the mmu_cr4 save it lost somehow. -v3: read back cr4 for 32bit too according to HPA -v4: use cpuid_level to check if we can read cr4 according to HPA don't touch mmu_cr4_features if CONFIG_HIBERNATION is defined from Yinghai -v5: reduce the using of CONFIG_HIBERNATION Bisected-and-tested-by: Michael Leun Signed-off-by: Stefano Stabellini Signed-off-by: Yinghai Lu --- arch/x86/include/asm/processor.h | 2 -- arch/x86/include/asm/suspend_32.h | 2 +- arch/x86/include/asm/suspend_64.h | 5 +---- arch/x86/kernel/setup.c | 16 +++++++++++++++- 4 files changed, 17 insertions(+), 8 deletions(-) Index: linux-2.6/arch/x86/kernel/setup.c =================================================================== --- linux-2.6.orig/arch/x86/kernel/setup.c +++ linux-2.6/arch/x86/kernel/setup.c @@ -212,12 +212,26 @@ struct cpuinfo_x86 boot_cpu_data __read_ EXPORT_SYMBOL(boot_cpu_data); #endif - +/* + * mmu_cr4_features has two purposes: + * a. head_32.S will access cr4 according if X86_CR4_PAE is set in it. + * b. store read back cr4 for hibernation + */ #if !defined(CONFIG_X86_PAE) || defined(CONFIG_X86_64) unsigned long mmu_cr4_features; #else unsigned long mmu_cr4_features = X86_CR4_PAE; #endif +#ifdef CONFIG_HIBERNATION +int arch_prepare_suspend(void) +{ + /* a CPU has CR4 iff it has CPUID --- hpa */ + if (boot_cpu_data.cpuid_level >= 0) + mmu_cr4_features = read_cr4(); + + return 0; +} +#endif /* Boot loader ID and version as integers, for the benefit of proc_dointvec */ int bootloader_type, bootloader_version; Index: linux-2.6/arch/x86/include/asm/processor.h =================================================================== --- linux-2.6.orig/arch/x86/include/asm/processor.h +++ linux-2.6/arch/x86/include/asm/processor.h @@ -601,7 +601,6 @@ static inline void set_in_cr4(unsigned l { unsigned long cr4; - mmu_cr4_features |= mask; cr4 = read_cr4(); cr4 |= mask; write_cr4(cr4); @@ -611,7 +610,6 @@ static inline void clear_in_cr4(unsigned { unsigned long cr4; - mmu_cr4_features &= ~mask; cr4 = read_cr4(); cr4 &= ~mask; write_cr4(cr4); Index: linux-2.6/arch/x86/include/asm/suspend_32.h =================================================================== --- linux-2.6.orig/arch/x86/include/asm/suspend_32.h +++ linux-2.6/arch/x86/include/asm/suspend_32.h @@ -9,7 +9,7 @@ #include #include -static inline int arch_prepare_suspend(void) { return 0; } +int arch_prepare_suspend(void); /* image of the saved processor state */ struct saved_context { Index: linux-2.6/arch/x86/include/asm/suspend_64.h =================================================================== --- linux-2.6.orig/arch/x86/include/asm/suspend_64.h +++ linux-2.6/arch/x86/include/asm/suspend_64.h @@ -9,10 +9,7 @@ #include #include -static inline int arch_prepare_suspend(void) -{ - return 0; -} +int arch_prepare_suspend(void); /* * Image of the saved processor state, used by the low level ACPI suspend to -- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/