From patchwork Sat Mar 14 07:34:05 2020 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Xiaoyao Li X-Patchwork-Id: 11438591 Return-Path: Received: from mail.kernel.org (pdx-korg-mail-1.web.codeaurora.org [172.30.200.123]) by pdx-korg-patchwork-2.web.codeaurora.org (Postfix) with ESMTP id BB0916CA for ; Sun, 15 Mar 2020 02:14:22 +0000 (UTC) Received: from vger.kernel.org (vger.kernel.org [209.132.180.67]) by mail.kernel.org (Postfix) with ESMTP id A4FEB2076E for ; Sun, 15 Mar 2020 02:14:22 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1727742AbgCOCNi (ORCPT ); Sat, 14 Mar 2020 22:13:38 -0400 Received: from mga14.intel.com ([192.55.52.115]:41895 "EHLO mga14.intel.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1727574AbgCOCNi (ORCPT ); Sat, 14 Mar 2020 22:13:38 -0400 IronPort-SDR: O/d525jcUaqxMiDOAY5ir7Zg2qZJBWPKz2sXqbkhieS17Y5e+Y455nnbMBFvQDK8nzXrEV+/Th w9yh+JhL0bVA== X-Amp-Result: SKIPPED(no attachment in message) X-Amp-File-Uploaded: False Received: from orsmga005.jf.intel.com ([10.7.209.41]) by fmsmga103.fm.intel.com with ESMTP/TLS/ECDHE-RSA-AES256-GCM-SHA384; 14 Mar 2020 00:51:51 -0700 IronPort-SDR: gsLoapCDm/VkIs7QUR8QRXIruDq/iLRJeTJMWr7AV+UhI/AJ8c4tU46gimiHbGp5eAkTEqYfPm v5Vfsz8hconA== X-ExtLoop1: 1 X-IronPort-AV: E=Sophos;i="5.70,551,1574150400"; d="scan'208";a="416537529" Received: from lxy-clx-4s.sh.intel.com ([10.239.43.160]) by orsmga005.jf.intel.com with ESMTP; 14 Mar 2020 00:51:47 -0700 From: Xiaoyao Li To: Thomas Gleixner , Ingo Molnar , Borislav Petkov , hpa@zytor.com, Paolo Bonzini , Sean Christopherson , Andy Lutomirski , tony.luck@intel.com Cc: peterz@infradead.org, fenghua.yu@intel.com, Arvind Sankar , Vitaly Kuznetsov , Jim Mattson , x86@kernel.org, kvm@vger.kernel.org, linux-kernel@vger.kernel.org, Xiaoyao Li Subject: [PATCH v4 01/10] x86/split_lock: Rework the initialization flow of split lock detection Date: Sat, 14 Mar 2020 15:34:05 +0800 Message-Id: <20200314073414.184213-2-xiaoyao.li@intel.com> X-Mailer: git-send-email 2.20.1 In-Reply-To: <20200314073414.184213-1-xiaoyao.li@intel.com> References: <20200314073414.184213-1-xiaoyao.li@intel.com> MIME-Version: 1.0 Sender: kvm-owner@vger.kernel.org Precedence: bulk List-ID: X-Mailing-List: kvm@vger.kernel.org Current initialization flow of split lock detection has following issues: 1. It assumes the initial value of MSR_TEST_CTRL.SPLIT_LOCK_DETECT to be zero. However, it's possible that BIOS/firmware has set it. 2. X86_FEATURE_SPLIT_LOCK_DETECT flag is unconditionally set even if there is a virtualization flaw that FMS indicates the existence while it's actually not supported. Introducing a new sld_state, "sld_not_exist", which is set as the default value. Only when sld_state != sld_not_exist, it goes to initialization flow. In initialization flow, it explicitly accesses MSR_TEST_CTRL and SPLIT_LOCK_DETECT bit to ensure there is no virtualization flaw. In detail, 1. sld_off, verify SPLIT_LOCK_DETECT bit can be cleared, and clear it; 2. sld_warn, verify SPLIT_LOCK_DETECT bit can be cleared and set, and set it; 3. sld_fatal, verify SPLIT_LOCK_DETECT bit can be set, and set it; Only when no MSR aceessing failure, can X86_FEATURE_SPLIT_LOCK_DETECT be set. Also, this can avoid using safe version RDMSR/WRMSR at runtime. Signed-off-by: Xiaoyao Li --- arch/x86/kernel/cpu/intel.c | 64 +++++++++++++++++++++++++------------ 1 file changed, 44 insertions(+), 20 deletions(-) diff --git a/arch/x86/kernel/cpu/intel.c b/arch/x86/kernel/cpu/intel.c index db3e745e5d47..064ba12defc8 100644 --- a/arch/x86/kernel/cpu/intel.c +++ b/arch/x86/kernel/cpu/intel.c @@ -34,17 +34,17 @@ #endif enum split_lock_detect_state { - sld_off = 0, + sld_not_exist = 0, + sld_off, sld_warn, sld_fatal, }; /* - * Default to sld_off because most systems do not support split lock detection * split_lock_setup() will switch this to sld_warn on systems that support * split lock detect, unless there is a command line override. */ -static enum split_lock_detect_state sld_state = sld_off; +static enum split_lock_detect_state sld_state = sld_not_exist; /* * Processors which have self-snooping capability can handle conflicting @@ -585,7 +585,7 @@ static void init_intel_misc_features(struct cpuinfo_x86 *c) wrmsrl(MSR_MISC_FEATURES_ENABLES, msr); } -static void split_lock_init(void); +static void split_lock_init(struct cpuinfo_x86 *c); static void init_intel(struct cpuinfo_x86 *c) { @@ -702,7 +702,8 @@ static void init_intel(struct cpuinfo_x86 *c) if (tsx_ctrl_state == TSX_CTRL_DISABLE) tsx_disable(); - split_lock_init(); + if (sld_state != sld_not_exist) + split_lock_init(c); } #ifdef CONFIG_X86_32 @@ -989,7 +990,6 @@ static void __init split_lock_setup(void) char arg[20]; int i, ret; - setup_force_cpu_cap(X86_FEATURE_SPLIT_LOCK_DETECT); sld_state = sld_warn; ret = cmdline_find_option(boot_command_line, "split_lock_detect", @@ -1015,6 +1015,8 @@ static void __init split_lock_setup(void) case sld_fatal: pr_info("sending SIGBUS on user-space split_locks\n"); break; + default: + break; } } @@ -1022,39 +1024,61 @@ static void __init split_lock_setup(void) * Locking is not required at the moment because only bit 29 of this * MSR is implemented and locking would not prevent that the operation * of one thread is immediately undone by the sibling thread. - * Use the "safe" versions of rdmsr/wrmsr here because although code - * checks CPUID and MSR bits to make sure the TEST_CTRL MSR should - * exist, there may be glitches in virtualization that leave a guest - * with an incorrect view of real h/w capabilities. */ -static bool __sld_msr_set(bool on) +static void __sld_msr_set(bool on) { u64 test_ctrl_val; - if (rdmsrl_safe(MSR_TEST_CTRL, &test_ctrl_val)) - return false; + rdmsrl(MSR_TEST_CTRL, test_ctrl_val); if (on) test_ctrl_val |= MSR_TEST_CTRL_SPLIT_LOCK_DETECT; else test_ctrl_val &= ~MSR_TEST_CTRL_SPLIT_LOCK_DETECT; - return !wrmsrl_safe(MSR_TEST_CTRL, test_ctrl_val); + wrmsrl(MSR_TEST_CTRL, test_ctrl_val); } -static void split_lock_init(void) +/* + * Use the "safe" versions of rdmsr/wrmsr here because although code + * checks CPUID and MSR bits to make sure the TEST_CTRL MSR should + * exist, there may be glitches in virtualization that leave a guest + * with an incorrect view of real h/w capabilities. + * If not msr_broken, then it needn't use "safe" version at runtime. + */ +static void split_lock_init(struct cpuinfo_x86 *c) { - if (sld_state == sld_off) - return; + u64 test_ctrl_val; - if (__sld_msr_set(true)) - return; + if (rdmsrl_safe(MSR_TEST_CTRL, &test_ctrl_val)) + goto msr_broken; + + switch (sld_state) { + case sld_off: + if (wrmsrl_safe(MSR_TEST_CTRL, test_ctrl_val & ~MSR_TEST_CTRL_SPLIT_LOCK_DETECT)) + goto msr_broken; + break; + case sld_warn: + if (wrmsrl_safe(MSR_TEST_CTRL, test_ctrl_val & ~MSR_TEST_CTRL_SPLIT_LOCK_DETECT)) + goto msr_broken; + fallthrough; + case sld_fatal: + if (wrmsrl_safe(MSR_TEST_CTRL, test_ctrl_val | MSR_TEST_CTRL_SPLIT_LOCK_DETECT)) + goto msr_broken; + break; + default: + break; + } + + set_cpu_cap(c, X86_FEATURE_SPLIT_LOCK_DETECT); + return; +msr_broken: /* * If this is anything other than the boot-cpu, you've done * funny things and you get to keep whatever pieces. */ - pr_warn("MSR fail -- disabled\n"); + pr_warn_once("MSR fail -- disabled\n"); sld_state = sld_off; } From patchwork Sat Mar 14 07:34:06 2020 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Xiaoyao Li X-Patchwork-Id: 11438593 Return-Path: Received: from mail.kernel.org (pdx-korg-mail-1.web.codeaurora.org [172.30.200.123]) by pdx-korg-patchwork-2.web.codeaurora.org (Postfix) with ESMTP id DE66B92C for ; Sun, 15 Mar 2020 02:14:28 +0000 (UTC) Received: from vger.kernel.org (vger.kernel.org [209.132.180.67]) by mail.kernel.org (Postfix) with ESMTP id C82DA2076E for ; Sun, 15 Mar 2020 02:14:28 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1727902AbgCOCOY (ORCPT ); Sat, 14 Mar 2020 22:14:24 -0400 Received: from mga14.intel.com ([192.55.52.115]:41895 "EHLO mga14.intel.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1727735AbgCOCNi (ORCPT ); Sat, 14 Mar 2020 22:13:38 -0400 IronPort-SDR: 8e8FIJ24Dzk4cpFyuHm3hy/CunTSuXhjbwEymkE68PzLvc/6v18+eC05K6luOSfws0pjOq+pbv T9HwKN8YQiCg== X-Amp-Result: SKIPPED(no attachment in message) X-Amp-File-Uploaded: False Received: from orsmga005.jf.intel.com ([10.7.209.41]) by fmsmga103.fm.intel.com with ESMTP/TLS/ECDHE-RSA-AES256-GCM-SHA384; 14 Mar 2020 00:51:55 -0700 IronPort-SDR: xFLMCsQOnVowPuCzZ86YJndKxfljywKlb7iR1Vj3Q0MbySU6B2XW7r8HjrUygPy8rzC5TSMUgn JercIOJwz/2Q== X-ExtLoop1: 1 X-IronPort-AV: E=Sophos;i="5.70,551,1574150400"; d="scan'208";a="416537546" Received: from lxy-clx-4s.sh.intel.com ([10.239.43.160]) by orsmga005.jf.intel.com with ESMTP; 14 Mar 2020 00:51:51 -0700 From: Xiaoyao Li To: Thomas Gleixner , Ingo Molnar , Borislav Petkov , hpa@zytor.com, Paolo Bonzini , Sean Christopherson , Andy Lutomirski , tony.luck@intel.com Cc: peterz@infradead.org, fenghua.yu@intel.com, Arvind Sankar , Vitaly Kuznetsov , Jim Mattson , x86@kernel.org, kvm@vger.kernel.org, linux-kernel@vger.kernel.org, Xiaoyao Li Subject: [PATCH v4 02/10] x86/split_lock: Avoid runtime reads of the TEST_CTRL MSR Date: Sat, 14 Mar 2020 15:34:06 +0800 Message-Id: <20200314073414.184213-3-xiaoyao.li@intel.com> X-Mailer: git-send-email 2.20.1 In-Reply-To: <20200314073414.184213-1-xiaoyao.li@intel.com> References: <20200314073414.184213-1-xiaoyao.li@intel.com> MIME-Version: 1.0 Sender: kvm-owner@vger.kernel.org Precedence: bulk List-ID: X-Mailing-List: kvm@vger.kernel.org In a context switch from a task that is detecting split locks to one that is not (or vice versa) we need to update the TEST_CTRL MSR. Currently this is done with the common sequence: read the MSR flip the bit write the MSR in order to avoid changing the value of any reserved bits in the MSR. Cache the value of the TEST_CTRL MSR when we read it during initialization so we can avoid an expensive RDMSR instruction during context switch. Suggested-by: Sean Christopherson Originally-by: Tony Luck Signed-off-by: Xiaoyao Li --- arch/x86/kernel/cpu/intel.c | 24 ++++++++++++++++-------- 1 file changed, 16 insertions(+), 8 deletions(-) diff --git a/arch/x86/kernel/cpu/intel.c b/arch/x86/kernel/cpu/intel.c index 064ba12defc8..4b3245035b5a 100644 --- a/arch/x86/kernel/cpu/intel.c +++ b/arch/x86/kernel/cpu/intel.c @@ -1020,6 +1020,14 @@ static void __init split_lock_setup(void) } } +/* + * Soft copy of MSR_TEST_CTRL initialized when we first read the + * MSR. Used at runtime to avoid using rdmsr again just to collect + * the reserved bits in the MSR. We assume reserved bits are the + * same on all CPUs. + */ +static u64 test_ctrl_val; + /* * Locking is not required at the moment because only bit 29 of this * MSR is implemented and locking would not prevent that the operation @@ -1027,16 +1035,14 @@ static void __init split_lock_setup(void) */ static void __sld_msr_set(bool on) { - u64 test_ctrl_val; - - rdmsrl(MSR_TEST_CTRL, test_ctrl_val); + u64 val = test_ctrl_val; if (on) - test_ctrl_val |= MSR_TEST_CTRL_SPLIT_LOCK_DETECT; + val |= MSR_TEST_CTRL_SPLIT_LOCK_DETECT; else - test_ctrl_val &= ~MSR_TEST_CTRL_SPLIT_LOCK_DETECT; + val &= ~MSR_TEST_CTRL_SPLIT_LOCK_DETECT; - wrmsrl(MSR_TEST_CTRL, test_ctrl_val); + wrmsrl(MSR_TEST_CTRL, val); } /* @@ -1048,11 +1054,13 @@ static void __sld_msr_set(bool on) */ static void split_lock_init(struct cpuinfo_x86 *c) { - u64 test_ctrl_val; + u64 val; - if (rdmsrl_safe(MSR_TEST_CTRL, &test_ctrl_val)) + if (rdmsrl_safe(MSR_TEST_CTRL, &val)) goto msr_broken; + test_ctrl_val = val; + switch (sld_state) { case sld_off: if (wrmsrl_safe(MSR_TEST_CTRL, test_ctrl_val & ~MSR_TEST_CTRL_SPLIT_LOCK_DETECT)) From patchwork Sat Mar 14 07:34:07 2020 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Xiaoyao Li X-Patchwork-Id: 11438575 Return-Path: Received: from mail.kernel.org (pdx-korg-mail-1.web.codeaurora.org [172.30.200.123]) by pdx-korg-patchwork-2.web.codeaurora.org (Postfix) with ESMTP id E684792C for ; Sun, 15 Mar 2020 02:13:48 +0000 (UTC) Received: from vger.kernel.org (vger.kernel.org [209.132.180.67]) by mail.kernel.org (Postfix) with ESMTP id D12962076D for ; Sun, 15 Mar 2020 02:13:48 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1727789AbgCOCNk (ORCPT ); Sat, 14 Mar 2020 22:13:40 -0400 Received: from mga14.intel.com ([192.55.52.115]:41895 "EHLO mga14.intel.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1727740AbgCOCNi (ORCPT ); Sat, 14 Mar 2020 22:13:38 -0400 IronPort-SDR: npGp+F9Wgxv+K+5FdaLHx6c5Q+wAKyWvtyTtvfJTUun6x16YUPTKjCU1Ld2kQmldZvAL/qmmEx B+j9u1bsgk1w== X-Amp-Result: SKIPPED(no attachment in message) X-Amp-File-Uploaded: False Received: from orsmga005.jf.intel.com ([10.7.209.41]) by fmsmga103.fm.intel.com with ESMTP/TLS/ECDHE-RSA-AES256-GCM-SHA384; 14 Mar 2020 00:52:00 -0700 IronPort-SDR: NqpETWF9PjHmdEwtXkD9iSSe6yB3S/7nAl9vxSwu7QqlFpCi5jPm2gjrrq1xiW4IW3rfMvdLvR NIABgARz6Yww== X-ExtLoop1: 1 X-IronPort-AV: E=Sophos;i="5.70,551,1574150400"; d="scan'208";a="416537556" Received: from lxy-clx-4s.sh.intel.com ([10.239.43.160]) by orsmga005.jf.intel.com with ESMTP; 14 Mar 2020 00:51:55 -0700 From: Xiaoyao Li To: Thomas Gleixner , Ingo Molnar , Borislav Petkov , hpa@zytor.com, Paolo Bonzini , Sean Christopherson , Andy Lutomirski , tony.luck@intel.com Cc: peterz@infradead.org, fenghua.yu@intel.com, Arvind Sankar , Vitaly Kuznetsov , Jim Mattson , x86@kernel.org, kvm@vger.kernel.org, linux-kernel@vger.kernel.org, Xiaoyao Li Subject: [PATCH v4 03/10] x86/split_lock: Re-define the kernel param option for split_lock_detect Date: Sat, 14 Mar 2020 15:34:07 +0800 Message-Id: <20200314073414.184213-4-xiaoyao.li@intel.com> X-Mailer: git-send-email 2.20.1 In-Reply-To: <20200314073414.184213-1-xiaoyao.li@intel.com> References: <20200314073414.184213-1-xiaoyao.li@intel.com> MIME-Version: 1.0 Sender: kvm-owner@vger.kernel.org Precedence: bulk List-ID: X-Mailing-List: kvm@vger.kernel.org Change sld_off to sld_disable, which means disabling feature split lock detection and it cannot be used in kernel nor can kvm expose it guest. Of course, the X86_FEATURE_SPLIT_LOCK_DETECT is not set. Add a new optioin sld_kvm_only, which means kernel turns split lock detection off, but kvm can expose it to guest. Signed-off-by: Xiaoyao Li --- .../admin-guide/kernel-parameters.txt | 5 ++++- arch/x86/kernel/cpu/intel.c | 22 ++++++++++++++----- 2 files changed, 21 insertions(+), 6 deletions(-) diff --git a/Documentation/admin-guide/kernel-parameters.txt b/Documentation/admin-guide/kernel-parameters.txt index 1ee2d1e6d89a..2b922061ff08 100644 --- a/Documentation/admin-guide/kernel-parameters.txt +++ b/Documentation/admin-guide/kernel-parameters.txt @@ -4666,7 +4666,10 @@ instructions that access data across cache line boundaries will result in an alignment check exception. - off - not enabled + disable - disabled, neither kernel nor kvm can use it. + + kvm_only - off in kernel but kvm can expose it to + guest for debug/testing scenario. warn - the kernel will emit rate limited warnings about applications triggering the #AC diff --git a/arch/x86/kernel/cpu/intel.c b/arch/x86/kernel/cpu/intel.c index 4b3245035b5a..3eeab717a0d0 100644 --- a/arch/x86/kernel/cpu/intel.c +++ b/arch/x86/kernel/cpu/intel.c @@ -35,7 +35,8 @@ enum split_lock_detect_state { sld_not_exist = 0, - sld_off, + sld_disable, + sld_kvm_only, sld_warn, sld_fatal, }; @@ -973,7 +974,8 @@ static const struct { const char *option; enum split_lock_detect_state state; } sld_options[] __initconst = { - { "off", sld_off }, + { "disable", sld_disable }, + { "kvm_only", sld_kvm_only }, { "warn", sld_warn }, { "fatal", sld_fatal }, }; @@ -1004,10 +1006,14 @@ static void __init split_lock_setup(void) } switch (sld_state) { - case sld_off: + case sld_disable: pr_info("disabled\n"); break; + case sld_kvm_only: + pr_info("off in kernel, but kvm can expose it to guest\n"); + break; + case sld_warn: pr_info("warning about user-space split_locks\n"); break; @@ -1062,7 +1068,13 @@ static void split_lock_init(struct cpuinfo_x86 *c) test_ctrl_val = val; switch (sld_state) { - case sld_off: + case sld_disable: + if (wrmsrl_safe(MSR_TEST_CTRL, test_ctrl_val & ~MSR_TEST_CTRL_SPLIT_LOCK_DETECT)) + goto msr_broken; + return; + case sld_kvm_only: + if (wrmsrl_safe(MSR_TEST_CTRL, test_ctrl_val | MSR_TEST_CTRL_SPLIT_LOCK_DETECT)) + goto msr_broken; if (wrmsrl_safe(MSR_TEST_CTRL, test_ctrl_val & ~MSR_TEST_CTRL_SPLIT_LOCK_DETECT)) goto msr_broken; break; @@ -1087,7 +1099,7 @@ static void split_lock_init(struct cpuinfo_x86 *c) * funny things and you get to keep whatever pieces. */ pr_warn_once("MSR fail -- disabled\n"); - sld_state = sld_off; + sld_state = sld_disable; } bool handle_user_split_lock(struct pt_regs *regs, long error_code) From patchwork Sat Mar 14 07:34:08 2020 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Xiaoyao Li X-Patchwork-Id: 11438589 Return-Path: Received: from mail.kernel.org (pdx-korg-mail-1.web.codeaurora.org [172.30.200.123]) by pdx-korg-patchwork-2.web.codeaurora.org (Postfix) with ESMTP id 4F50B92C for ; Sun, 15 Mar 2020 02:14:20 +0000 (UTC) Received: from vger.kernel.org (vger.kernel.org [209.132.180.67]) by mail.kernel.org (Postfix) with ESMTP id 396C82076E for ; Sun, 15 Mar 2020 02:14:20 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1727873AbgCOCOQ (ORCPT ); Sat, 14 Mar 2020 22:14:16 -0400 Received: from mga14.intel.com ([192.55.52.115]:41896 "EHLO mga14.intel.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1727744AbgCOCNi (ORCPT ); Sat, 14 Mar 2020 22:13:38 -0400 IronPort-SDR: ec91g7YqAudwWfp540QIr9XBcc6G58vomOvum3iIYQCHrz8CyxpLu9CnFPaN0M5l46M4ey16uq IPjZD04o5JMw== X-Amp-Result: SKIPPED(no attachment in message) X-Amp-File-Uploaded: False Received: from orsmga005.jf.intel.com ([10.7.209.41]) by fmsmga103.fm.intel.com with ESMTP/TLS/ECDHE-RSA-AES256-GCM-SHA384; 14 Mar 2020 00:52:04 -0700 IronPort-SDR: +7QhIa6OG5TdA+1pLZPBQIvwnIurqmL9c+HZwZqJV7JzM6sbE3EvuEnyngT79jQWYeu8C+nlDn Lv43KW7SkTjQ== X-ExtLoop1: 1 X-IronPort-AV: E=Sophos;i="5.70,551,1574150400"; d="scan'208";a="416537579" Received: from lxy-clx-4s.sh.intel.com ([10.239.43.160]) by orsmga005.jf.intel.com with ESMTP; 14 Mar 2020 00:52:00 -0700 From: Xiaoyao Li To: Thomas Gleixner , Ingo Molnar , Borislav Petkov , hpa@zytor.com, Paolo Bonzini , Sean Christopherson , Andy Lutomirski , tony.luck@intel.com Cc: peterz@infradead.org, fenghua.yu@intel.com, Arvind Sankar , Vitaly Kuznetsov , Jim Mattson , x86@kernel.org, kvm@vger.kernel.org, linux-kernel@vger.kernel.org, Xiaoyao Li Subject: [PATCH v4 04/10] x86/split_lock: Export handle_user_split_lock() Date: Sat, 14 Mar 2020 15:34:08 +0800 Message-Id: <20200314073414.184213-5-xiaoyao.li@intel.com> X-Mailer: git-send-email 2.20.1 In-Reply-To: <20200314073414.184213-1-xiaoyao.li@intel.com> References: <20200314073414.184213-1-xiaoyao.li@intel.com> MIME-Version: 1.0 Sender: kvm-owner@vger.kernel.org Precedence: bulk List-ID: X-Mailing-List: kvm@vger.kernel.org In the future, KVM will use handle_user_split_lock() to handle #AC caused by split lock in guest. Due to the fact that KVM doesn't have a @regs context and will pre-check EFLASG.AC, move the EFLAGS.AC check to do_alignment_check(). Suggested-by: Sean Christopherson Signed-off-by: Xiaoyao Li --- arch/x86/include/asm/cpu.h | 4 ++-- arch/x86/kernel/cpu/intel.c | 7 ++++--- arch/x86/kernel/traps.c | 2 +- 3 files changed, 7 insertions(+), 6 deletions(-) diff --git a/arch/x86/include/asm/cpu.h b/arch/x86/include/asm/cpu.h index ff6f3ca649b3..ff567afa6ee1 100644 --- a/arch/x86/include/asm/cpu.h +++ b/arch/x86/include/asm/cpu.h @@ -43,11 +43,11 @@ unsigned int x86_stepping(unsigned int sig); #ifdef CONFIG_CPU_SUP_INTEL extern void __init cpu_set_core_cap_bits(struct cpuinfo_x86 *c); extern void switch_to_sld(unsigned long tifn); -extern bool handle_user_split_lock(struct pt_regs *regs, long error_code); +extern bool handle_user_split_lock(unsigned long ip); #else static inline void __init cpu_set_core_cap_bits(struct cpuinfo_x86 *c) {} static inline void switch_to_sld(unsigned long tifn) {} -static inline bool handle_user_split_lock(struct pt_regs *regs, long error_code) +static inline bool handle_user_split_lock(unsigned long ip) { return false; } diff --git a/arch/x86/kernel/cpu/intel.c b/arch/x86/kernel/cpu/intel.c index 3eeab717a0d0..c401d174c8db 100644 --- a/arch/x86/kernel/cpu/intel.c +++ b/arch/x86/kernel/cpu/intel.c @@ -1102,13 +1102,13 @@ static void split_lock_init(struct cpuinfo_x86 *c) sld_state = sld_disable; } -bool handle_user_split_lock(struct pt_regs *regs, long error_code) +bool handle_user_split_lock(unsigned long ip) { - if ((regs->flags & X86_EFLAGS_AC) || sld_state == sld_fatal) + if (sld_state == sld_fatal) return false; pr_warn_ratelimited("#AC: %s/%d took a split_lock trap at address: 0x%lx\n", - current->comm, current->pid, regs->ip); + current->comm, current->pid, ip); /* * Disable the split lock detection for this task so it can make @@ -1119,6 +1119,7 @@ bool handle_user_split_lock(struct pt_regs *regs, long error_code) set_tsk_thread_flag(current, TIF_SLD); return true; } +EXPORT_SYMBOL_GPL(handle_user_split_lock); /* * This function is called only when switching between tasks with diff --git a/arch/x86/kernel/traps.c b/arch/x86/kernel/traps.c index 0ef5befaed7d..407ff9be610f 100644 --- a/arch/x86/kernel/traps.c +++ b/arch/x86/kernel/traps.c @@ -304,7 +304,7 @@ dotraplinkage void do_alignment_check(struct pt_regs *regs, long error_code) local_irq_enable(); - if (handle_user_split_lock(regs, error_code)) + if (!(regs->flags & X86_EFLAGS_AC) && handle_user_split_lock(regs->ip)) return; do_trap(X86_TRAP_AC, SIGBUS, "alignment check", regs, From patchwork Sat Mar 14 07:34:09 2020 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Xiaoyao Li X-Patchwork-Id: 11438585 Return-Path: Received: from mail.kernel.org (pdx-korg-mail-1.web.codeaurora.org [172.30.200.123]) by pdx-korg-patchwork-2.web.codeaurora.org (Postfix) with ESMTP id CA0196CA for ; Sun, 15 Mar 2020 02:14:10 +0000 (UTC) Received: from vger.kernel.org (vger.kernel.org [209.132.180.67]) by mail.kernel.org (Postfix) with ESMTP id B44D72076D for ; Sun, 15 Mar 2020 02:14:10 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1727764AbgCOCNj (ORCPT ); Sat, 14 Mar 2020 22:13:39 -0400 Received: from mga14.intel.com ([192.55.52.115]:41896 "EHLO mga14.intel.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1727751AbgCOCNi (ORCPT ); Sat, 14 Mar 2020 22:13:38 -0400 IronPort-SDR: qPqE0DsQzM48T3slRjEdVhjYWsJlNRUwp+vrd0E/xa7qe6TZJZRWNsKFZh6+roKkJUEB2dAiww BA5EupNJTUYQ== X-Amp-Result: SKIPPED(no attachment in message) X-Amp-File-Uploaded: False Received: from orsmga005.jf.intel.com ([10.7.209.41]) by fmsmga103.fm.intel.com with ESMTP/TLS/ECDHE-RSA-AES256-GCM-SHA384; 14 Mar 2020 00:52:08 -0700 IronPort-SDR: W+uTITCYwRT4BUq7itpYFcStZgrU00B48B5pMT0zv7x7LiZwh44JOXWVD7LE56tMfqsvI9Mcno mPiCzBtgzs4A== X-ExtLoop1: 1 X-IronPort-AV: E=Sophos;i="5.70,551,1574150400"; d="scan'208";a="416537597" Received: from lxy-clx-4s.sh.intel.com ([10.239.43.160]) by orsmga005.jf.intel.com with ESMTP; 14 Mar 2020 00:52:04 -0700 From: Xiaoyao Li To: Thomas Gleixner , Ingo Molnar , Borislav Petkov , hpa@zytor.com, Paolo Bonzini , Sean Christopherson , Andy Lutomirski , tony.luck@intel.com Cc: peterz@infradead.org, fenghua.yu@intel.com, Arvind Sankar , Vitaly Kuznetsov , Jim Mattson , x86@kernel.org, kvm@vger.kernel.org, linux-kernel@vger.kernel.org, Xiaoyao Li Subject: [PATCH v4 05/10] x86/split_lock: Add and export several functions for KVM Date: Sat, 14 Mar 2020 15:34:09 +0800 Message-Id: <20200314073414.184213-6-xiaoyao.li@intel.com> X-Mailer: git-send-email 2.20.1 In-Reply-To: <20200314073414.184213-1-xiaoyao.li@intel.com> References: <20200314073414.184213-1-xiaoyao.li@intel.com> MIME-Version: 1.0 Sender: kvm-owner@vger.kernel.org Precedence: bulk List-ID: X-Mailing-List: kvm@vger.kernel.org KVM will use split_lock_detect_disabled() and split_lock_detect_on() in vmx_vcpu_run() to check whether split lock detect can be exposed to guest and whether host has turned it on. Make them static inline to avoid the extra CALL+RET in that path. sld_msr_set() will be used when switching from/to guest. Signed-off-by: Xiaoyao Li --- arch/x86/include/asm/cpu.h | 23 +++++++++++++++++++++++ arch/x86/kernel/cpu/intel.c | 17 ++++++++--------- 2 files changed, 31 insertions(+), 9 deletions(-) diff --git a/arch/x86/include/asm/cpu.h b/arch/x86/include/asm/cpu.h index ff567afa6ee1..2e17315b1fed 100644 --- a/arch/x86/include/asm/cpu.h +++ b/arch/x86/include/asm/cpu.h @@ -41,15 +41,38 @@ unsigned int x86_family(unsigned int sig); unsigned int x86_model(unsigned int sig); unsigned int x86_stepping(unsigned int sig); #ifdef CONFIG_CPU_SUP_INTEL +enum split_lock_detect_state { + sld_not_exist = 0, + sld_disable, + sld_kvm_only, + sld_warn, + sld_fatal, +}; +extern enum split_lock_detect_state sld_state; + +static inline bool split_lock_detect_on(void) +{ + return (sld_state == sld_warn) || (sld_state == sld_fatal); +} + +static inline bool split_lock_detect_disabled(void) +{ + return sld_state == sld_disable; +} + extern void __init cpu_set_core_cap_bits(struct cpuinfo_x86 *c); extern void switch_to_sld(unsigned long tifn); extern bool handle_user_split_lock(unsigned long ip); +extern void sld_msr_set(bool on); #else +static inline bool split_lock_detect_on(void) { return false; } +static inline bool split_lock_detect_disabled(void) { return true; } static inline void __init cpu_set_core_cap_bits(struct cpuinfo_x86 *c) {} static inline void switch_to_sld(unsigned long tifn) {} static inline bool handle_user_split_lock(unsigned long ip) { return false; } +static inline void sld_msr_set(bool on) {} #endif #endif /* _ASM_X86_CPU_H */ diff --git a/arch/x86/kernel/cpu/intel.c b/arch/x86/kernel/cpu/intel.c index c401d174c8db..8bfe8b07e06e 100644 --- a/arch/x86/kernel/cpu/intel.c +++ b/arch/x86/kernel/cpu/intel.c @@ -33,19 +33,12 @@ #include #endif -enum split_lock_detect_state { - sld_not_exist = 0, - sld_disable, - sld_kvm_only, - sld_warn, - sld_fatal, -}; - /* * split_lock_setup() will switch this to sld_warn on systems that support * split lock detect, unless there is a command line override. */ -static enum split_lock_detect_state sld_state = sld_not_exist; +enum split_lock_detect_state sld_state = sld_not_exist; +EXPORT_SYMBOL_GPL(sld_state); /* * Processors which have self-snooping capability can handle conflicting @@ -1121,6 +1114,12 @@ bool handle_user_split_lock(unsigned long ip) } EXPORT_SYMBOL_GPL(handle_user_split_lock); +void sld_msr_set(bool on) +{ + __sld_msr_set(on); +} +EXPORT_SYMBOL_GPL(sld_msr_set); + /* * This function is called only when switching between tasks with * different split-lock detection modes. It sets the MSR for the From patchwork Sat Mar 14 07:34:10 2020 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Xiaoyao Li X-Patchwork-Id: 11438577 Return-Path: Received: from mail.kernel.org (pdx-korg-mail-1.web.codeaurora.org [172.30.200.123]) by pdx-korg-patchwork-2.web.codeaurora.org (Postfix) with ESMTP id 0D89792C for ; Sun, 15 Mar 2020 02:13:51 +0000 (UTC) Received: from vger.kernel.org (vger.kernel.org [209.132.180.67]) by mail.kernel.org (Postfix) with ESMTP id EBCB72076A for ; Sun, 15 Mar 2020 02:13:50 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1727775AbgCOCNk (ORCPT ); Sat, 14 Mar 2020 22:13:40 -0400 Received: from mga14.intel.com ([192.55.52.115]:41895 "EHLO mga14.intel.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1727574AbgCOCNj (ORCPT ); Sat, 14 Mar 2020 22:13:39 -0400 IronPort-SDR: 8w94JfkDG2xmB8Zj3qCbdxkKzPnQN04wZn6ZS/xI3IsK8viv/EfahGDqaeNSuSkNtHez/ghzQQ /+dbxaWPK9pQ== X-Amp-Result: SKIPPED(no attachment in message) X-Amp-File-Uploaded: False Received: from orsmga005.jf.intel.com ([10.7.209.41]) by fmsmga103.fm.intel.com with ESMTP/TLS/ECDHE-RSA-AES256-GCM-SHA384; 14 Mar 2020 00:52:12 -0700 IronPort-SDR: 1kMrL2DKDmDEcoNhUZ/5FfRaCFdM0wYRIxRufPV34/znYIGFPlNv/tH9OuPwI1r/nKcOtdJx9h kAA2ssd+TzMg== X-ExtLoop1: 1 X-IronPort-AV: E=Sophos;i="5.70,551,1574150400"; d="scan'208";a="416537608" Received: from lxy-clx-4s.sh.intel.com ([10.239.43.160]) by orsmga005.jf.intel.com with ESMTP; 14 Mar 2020 00:52:08 -0700 From: Xiaoyao Li To: Thomas Gleixner , Ingo Molnar , Borislav Petkov , hpa@zytor.com, Paolo Bonzini , Sean Christopherson , Andy Lutomirski , tony.luck@intel.com Cc: peterz@infradead.org, fenghua.yu@intel.com, Arvind Sankar , Vitaly Kuznetsov , Jim Mattson , x86@kernel.org, kvm@vger.kernel.org, linux-kernel@vger.kernel.org, Xiaoyao Li Subject: [PATCH v4 06/10] kvm: x86: Emulate split-lock access as a write Date: Sat, 14 Mar 2020 15:34:10 +0800 Message-Id: <20200314073414.184213-7-xiaoyao.li@intel.com> X-Mailer: git-send-email 2.20.1 In-Reply-To: <20200314073414.184213-1-xiaoyao.li@intel.com> References: <20200314073414.184213-1-xiaoyao.li@intel.com> MIME-Version: 1.0 Sender: kvm-owner@vger.kernel.org Precedence: bulk List-ID: X-Mailing-List: kvm@vger.kernel.org If split lock detect is enabled (warn/fatal), #AC handler calls die() when split lock happens in kernel. Malicous guest can exploit the KVM emulator to trigger split lock #AC in kernel[1]. So just emulating the access as a write if it's a split-lock access (the same as access spans page) to avoid malicious attacking kernel. More discussion can be found [2][3]. [1] https://lore.kernel.org/lkml/8c5b11c9-58df-38e7-a514-dc12d687b198@redhat.com/ [2] https://lkml.kernel.org/r/20200131200134.GD18946@linux.intel.com [3] https://lkml.kernel.org/r/20200227001117.GX9940@linux.intel.com Suggested-by: Sean Christopherson Signed-off-by: Xiaoyao Li --- arch/x86/kvm/x86.c | 7 ++++++- 1 file changed, 6 insertions(+), 1 deletion(-) diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c index 5de200663f51..1a0e6c0b1b39 100644 --- a/arch/x86/kvm/x86.c +++ b/arch/x86/kvm/x86.c @@ -5873,6 +5873,7 @@ static int emulator_cmpxchg_emulated(struct x86_emulate_ctxt *ctxt, { struct kvm_host_map map; struct kvm_vcpu *vcpu = emul_to_vcpu(ctxt); + u64 page_line_mask = PAGE_MASK; gpa_t gpa; char *kaddr; bool exchanged; @@ -5887,7 +5888,11 @@ static int emulator_cmpxchg_emulated(struct x86_emulate_ctxt *ctxt, (gpa & PAGE_MASK) == APIC_DEFAULT_PHYS_BASE) goto emul_write; - if (((gpa + bytes - 1) & PAGE_MASK) != (gpa & PAGE_MASK)) + if (split_lock_detect_on()) + page_line_mask = ~(cache_line_size() - 1); + + /* when write spans page or spans cache when SLD enabled */ + if (((gpa + bytes - 1) & page_line_mask) != (gpa & page_line_mask)) goto emul_write; if (kvm_vcpu_map(vcpu, gpa_to_gfn(gpa), &map)) From patchwork Sat Mar 14 07:34:11 2020 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Xiaoyao Li X-Patchwork-Id: 11438587 Return-Path: Received: from mail.kernel.org (pdx-korg-mail-1.web.codeaurora.org [172.30.200.123]) by pdx-korg-patchwork-2.web.codeaurora.org (Postfix) with ESMTP id 8BFA592C for ; Sun, 15 Mar 2020 02:14:12 +0000 (UTC) Received: from vger.kernel.org (vger.kernel.org [209.132.180.67]) by mail.kernel.org (Postfix) with ESMTP id 76D1920786 for ; Sun, 15 Mar 2020 02:14:12 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1727855AbgCOCOL (ORCPT ); Sat, 14 Mar 2020 22:14:11 -0400 Received: from mga14.intel.com ([192.55.52.115]:41896 "EHLO mga14.intel.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1727754AbgCOCNj (ORCPT ); Sat, 14 Mar 2020 22:13:39 -0400 IronPort-SDR: 7Dx4jeO+q7GNQWymEQd0AmGCAljbX3G3jr/R66WE6XbhpgazNs6dFJ9x9+3nV+NdOLcBzCrJcN 0AxC0rfZa5FA== X-Amp-Result: SKIPPED(no attachment in message) X-Amp-File-Uploaded: False Received: from orsmga005.jf.intel.com ([10.7.209.41]) by fmsmga103.fm.intel.com with ESMTP/TLS/ECDHE-RSA-AES256-GCM-SHA384; 14 Mar 2020 00:52:17 -0700 IronPort-SDR: sIDcx8X+9yNrEo0qlSTkO5PF8Ib4uZZZPKiUy7hNOCsuIlyf0SaXWZ+yJnUvsYOjZW35LHnX9T N9I1lsUmJncA== X-ExtLoop1: 1 X-IronPort-AV: E=Sophos;i="5.70,551,1574150400"; d="scan'208";a="416537619" Received: from lxy-clx-4s.sh.intel.com ([10.239.43.160]) by orsmga005.jf.intel.com with ESMTP; 14 Mar 2020 00:52:12 -0700 From: Xiaoyao Li To: Thomas Gleixner , Ingo Molnar , Borislav Petkov , hpa@zytor.com, Paolo Bonzini , Sean Christopherson , Andy Lutomirski , tony.luck@intel.com Cc: peterz@infradead.org, fenghua.yu@intel.com, Arvind Sankar , Vitaly Kuznetsov , Jim Mattson , x86@kernel.org, kvm@vger.kernel.org, linux-kernel@vger.kernel.org, Xiaoyao Li Subject: [PATCH v4 07/10] kvm: vmx: Extend VMX's #AC interceptor to handle split lock #AC happens in guest Date: Sat, 14 Mar 2020 15:34:11 +0800 Message-Id: <20200314073414.184213-8-xiaoyao.li@intel.com> X-Mailer: git-send-email 2.20.1 In-Reply-To: <20200314073414.184213-1-xiaoyao.li@intel.com> References: <20200314073414.184213-1-xiaoyao.li@intel.com> MIME-Version: 1.0 Sender: kvm-owner@vger.kernel.org Precedence: bulk List-ID: X-Mailing-List: kvm@vger.kernel.org There are two types of #AC can be generated in Intel CPUs: 1. legacy alignment check #AC; 2. split lock #AC; Legacy alignment check #AC can be injected to guest if guest has enabled alignemnet check. when host enables split lock detectin, i.e., sld_warn or sld_fatal, there will be an unexpected #AC in guest and intercepted by KVM because KVM doesn't virtualize this feature to guest and hardware value of MSR_TEST_CTRL.SLD bit stays unchanged when vcpu is running. To handle this unexpected #AC, treat guest just like host usermode that calling handle_user_split_lock(): - If host is sld_warn, it warns and set TIF_SLD so that __switch_to_xtra() does the MSR_TEST_CTRL.SLD bit switching when control transfer to/from this vcpu. - If host is sld_fatal, forward #AC to userspace, the similar as sending SIGBUS. Suggested-by: Sean Christopherson Signed-off-by: Xiaoyao Li --- arch/x86/kvm/vmx/vmx.c | 30 +++++++++++++++++++++++++++--- 1 file changed, 27 insertions(+), 3 deletions(-) diff --git a/arch/x86/kvm/vmx/vmx.c b/arch/x86/kvm/vmx/vmx.c index 40b1e6138cd5..3fb132ad489d 100644 --- a/arch/x86/kvm/vmx/vmx.c +++ b/arch/x86/kvm/vmx/vmx.c @@ -4609,6 +4609,12 @@ static int handle_machine_check(struct kvm_vcpu *vcpu) return 1; } +static inline bool guest_cpu_alignment_check_enabled(struct kvm_vcpu *vcpu) +{ + return vmx_get_cpl(vcpu) == 3 && kvm_read_cr0_bits(vcpu, X86_CR0_AM) && + (kvm_get_rflags(vcpu) & X86_EFLAGS_AC); +} + static int handle_exception_nmi(struct kvm_vcpu *vcpu) { struct vcpu_vmx *vmx = to_vmx(vcpu); @@ -4674,9 +4680,6 @@ static int handle_exception_nmi(struct kvm_vcpu *vcpu) return handle_rmode_exception(vcpu, ex_no, error_code); switch (ex_no) { - case AC_VECTOR: - kvm_queue_exception_e(vcpu, AC_VECTOR, error_code); - return 1; case DB_VECTOR: dr6 = vmcs_readl(EXIT_QUALIFICATION); if (!(vcpu->guest_debug & @@ -4705,6 +4708,27 @@ static int handle_exception_nmi(struct kvm_vcpu *vcpu) kvm_run->debug.arch.pc = vmcs_readl(GUEST_CS_BASE) + rip; kvm_run->debug.arch.exception = ex_no; break; + case AC_VECTOR: + /* + * Reflect #AC to the guest if it's expecting the #AC, i.e. has + * legacy alignment check enabled. Pre-check host split lock + * support to avoid the VMREADs needed to check legacy #AC, + * i.e. reflect the #AC if the only possible source is legacy + * alignment checks. + */ + if (!split_lock_detect_on() || + guest_cpu_alignment_check_enabled(vcpu)) { + kvm_queue_exception_e(vcpu, AC_VECTOR, error_code); + return 1; + } + + /* + * Forward the #AC to userspace if kernel policy does not allow + * temporarily disabling split lock detection. + */ + if (handle_user_split_lock(kvm_rip_read(vcpu))) + return 1; + fallthrough; default: kvm_run->exit_reason = KVM_EXIT_EXCEPTION; kvm_run->ex.exception = ex_no; From patchwork Sat Mar 14 07:34:12 2020 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Xiaoyao Li X-Patchwork-Id: 11438579 Return-Path: Received: from mail.kernel.org (pdx-korg-mail-1.web.codeaurora.org [172.30.200.123]) by pdx-korg-patchwork-2.web.codeaurora.org (Postfix) with ESMTP id 2D91F6CA for ; Sun, 15 Mar 2020 02:13:51 +0000 (UTC) Received: from vger.kernel.org (vger.kernel.org [209.132.180.67]) by mail.kernel.org (Postfix) with ESMTP id 17A392076A for ; Sun, 15 Mar 2020 02:13:51 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1727813AbgCOCNu (ORCPT ); Sat, 14 Mar 2020 22:13:50 -0400 Received: from mga14.intel.com ([192.55.52.115]:41895 "EHLO mga14.intel.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1727756AbgCOCNk (ORCPT ); Sat, 14 Mar 2020 22:13:40 -0400 IronPort-SDR: kEZzYNe9dEsDwGvpKmxCKUAqdMS0HGtWD4ooH/WYc1g0A+ayHNRu7/J8I4KExy2IzJF230zXDB NjJGTQ8VLjZw== X-Amp-Result: SKIPPED(no attachment in message) X-Amp-File-Uploaded: False Received: from orsmga005.jf.intel.com ([10.7.209.41]) by fmsmga103.fm.intel.com with ESMTP/TLS/ECDHE-RSA-AES256-GCM-SHA384; 14 Mar 2020 00:52:21 -0700 IronPort-SDR: jnEP2ylUcG7iujopFUhnByVTJqDQk9h9uOQj+zSglOw2zIzsYCKFObOhVAIGwUPjWHwpj0nxMX ePt88Nlk3daA== X-ExtLoop1: 1 X-IronPort-AV: E=Sophos;i="5.70,551,1574150400"; d="scan'208";a="416537634" Received: from lxy-clx-4s.sh.intel.com ([10.239.43.160]) by orsmga005.jf.intel.com with ESMTP; 14 Mar 2020 00:52:17 -0700 From: Xiaoyao Li To: Thomas Gleixner , Ingo Molnar , Borislav Petkov , hpa@zytor.com, Paolo Bonzini , Sean Christopherson , Andy Lutomirski , tony.luck@intel.com Cc: peterz@infradead.org, fenghua.yu@intel.com, Arvind Sankar , Vitaly Kuznetsov , Jim Mattson , x86@kernel.org, kvm@vger.kernel.org, linux-kernel@vger.kernel.org, Xiaoyao Li Subject: [PATCH v4 08/10] kvm: x86: Emulate MSR IA32_CORE_CAPABILITIES Date: Sat, 14 Mar 2020 15:34:12 +0800 Message-Id: <20200314073414.184213-9-xiaoyao.li@intel.com> X-Mailer: git-send-email 2.20.1 In-Reply-To: <20200314073414.184213-1-xiaoyao.li@intel.com> References: <20200314073414.184213-1-xiaoyao.li@intel.com> MIME-Version: 1.0 Sender: kvm-owner@vger.kernel.org Precedence: bulk List-ID: X-Mailing-List: kvm@vger.kernel.org Emulate MSR_IA32_CORE_CAPABILITIES in software and unconditionally advertise its support to userspace. Like MSR_IA32_ARCH_CAPABILITIES, it is a feature-enumerating MSR and can be fully emulated regardless of hardware support. Existence of CORE_CAPABILITIES is enumerated via CPUID.(EAX=7H,ECX=0):EDX[30]. Note, support for individual features enumerated via CORE_CAPABILITIES, e.g., split lock detection, will be added in future patches. Signed-off-by: Xiaoyao Li --- arch/x86/include/asm/kvm_host.h | 1 + arch/x86/kvm/cpuid.c | 7 ++++--- arch/x86/kvm/x86.c | 22 ++++++++++++++++++++++ 3 files changed, 27 insertions(+), 3 deletions(-) diff --git a/arch/x86/include/asm/kvm_host.h b/arch/x86/include/asm/kvm_host.h index 98959e8cd448..d538c9f51a09 100644 --- a/arch/x86/include/asm/kvm_host.h +++ b/arch/x86/include/asm/kvm_host.h @@ -593,6 +593,7 @@ struct kvm_vcpu_arch { u64 ia32_xss; u64 microcode_version; u64 arch_capabilities; + u64 core_capabilities; /* * Paging state of the vcpu diff --git a/arch/x86/kvm/cpuid.c b/arch/x86/kvm/cpuid.c index b1c469446b07..344cd605ecaa 100644 --- a/arch/x86/kvm/cpuid.c +++ b/arch/x86/kvm/cpuid.c @@ -374,7 +374,7 @@ static inline void do_cpuid_7_mask(struct kvm_cpuid_entry2 *entry, int index) const u32 kvm_cpuid_7_0_edx_x86_features = F(AVX512_4VNNIW) | F(AVX512_4FMAPS) | F(SPEC_CTRL) | F(SPEC_CTRL_SSBD) | F(ARCH_CAPABILITIES) | F(INTEL_STIBP) | - F(MD_CLEAR); + F(MD_CLEAR) | F(CORE_CAPABILITIES); /* cpuid 7.1.eax */ const u32 kvm_cpuid_7_1_eax_x86_features = @@ -409,10 +409,11 @@ static inline void do_cpuid_7_mask(struct kvm_cpuid_entry2 *entry, int index) boot_cpu_has(X86_FEATURE_AMD_SSBD)) entry->edx |= F(SPEC_CTRL_SSBD); /* - * We emulate ARCH_CAPABILITIES in software even - * if the host doesn't support it. + * ARCH_CAPABILITIES and CORE_CAPABILITIES are emulated in + * software regardless of host support. */ entry->edx |= F(ARCH_CAPABILITIES); + entry->edx |= F(CORE_CAPABILITIES); break; case 1: entry->eax &= kvm_cpuid_7_1_eax_x86_features; diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c index 1a0e6c0b1b39..72d4bfea8864 100644 --- a/arch/x86/kvm/x86.c +++ b/arch/x86/kvm/x86.c @@ -1221,6 +1221,7 @@ static const u32 emulated_msrs_all[] = { MSR_IA32_TSC_ADJUST, MSR_IA32_TSCDEADLINE, MSR_IA32_ARCH_CAPABILITIES, + MSR_IA32_CORE_CAPS, MSR_IA32_MISC_ENABLE, MSR_IA32_MCG_STATUS, MSR_IA32_MCG_CTL, @@ -1287,6 +1288,7 @@ static const u32 msr_based_features_all[] = { MSR_F10H_DECFG, MSR_IA32_UCODE_REV, MSR_IA32_ARCH_CAPABILITIES, + MSR_IA32_CORE_CAPS, }; static u32 msr_based_features[ARRAY_SIZE(msr_based_features_all)]; @@ -1340,12 +1342,20 @@ static u64 kvm_get_arch_capabilities(void) return data; } +static u64 kvm_get_core_capabilities(void) +{ + return 0; +} + static int kvm_get_msr_feature(struct kvm_msr_entry *msr) { switch (msr->index) { case MSR_IA32_ARCH_CAPABILITIES: msr->data = kvm_get_arch_capabilities(); break; + case MSR_IA32_CORE_CAPS: + msr->data = kvm_get_core_capabilities(); + break; case MSR_IA32_UCODE_REV: rdmsrl_safe(msr->index, &msr->data); break; @@ -2718,6 +2728,11 @@ int kvm_set_msr_common(struct kvm_vcpu *vcpu, struct msr_data *msr_info) return 1; vcpu->arch.arch_capabilities = data; break; + case MSR_IA32_CORE_CAPS: + if (!msr_info->host_initiated) + return 1; + vcpu->arch.core_capabilities = data; + break; case MSR_EFER: return set_efer(vcpu, msr_info); case MSR_K7_HWCR: @@ -3046,6 +3061,12 @@ int kvm_get_msr_common(struct kvm_vcpu *vcpu, struct msr_data *msr_info) return 1; msr_info->data = vcpu->arch.arch_capabilities; break; + case MSR_IA32_CORE_CAPS: + if (!msr_info->host_initiated && + !guest_cpuid_has(vcpu, X86_FEATURE_CORE_CAPABILITIES)) + return 1; + msr_info->data = vcpu->arch.core_capabilities; + break; case MSR_IA32_POWER_CTL: msr_info->data = vcpu->arch.msr_ia32_power_ctl; break; @@ -9348,6 +9369,7 @@ int kvm_arch_vcpu_create(struct kvm_vcpu *vcpu) goto free_guest_fpu; vcpu->arch.arch_capabilities = kvm_get_arch_capabilities(); + vcpu->arch.core_capabilities = kvm_get_core_capabilities(); vcpu->arch.msr_platform_info = MSR_PLATFORM_INFO_CPUID_FAULT; kvm_vcpu_mtrr_init(vcpu); vcpu_load(vcpu); From patchwork Sat Mar 14 07:34:13 2020 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Xiaoyao Li X-Patchwork-Id: 11438583 Return-Path: Received: from mail.kernel.org (pdx-korg-mail-1.web.codeaurora.org [172.30.200.123]) by pdx-korg-patchwork-2.web.codeaurora.org (Postfix) with ESMTP id C09B16CA for ; Sun, 15 Mar 2020 02:14:06 +0000 (UTC) Received: from vger.kernel.org (vger.kernel.org [209.132.180.67]) by mail.kernel.org (Postfix) with ESMTP id AA80F2076A for ; Sun, 15 Mar 2020 02:14:06 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1727768AbgCOCOD (ORCPT ); Sat, 14 Mar 2020 22:14:03 -0400 Received: from mga14.intel.com ([192.55.52.115]:41898 "EHLO mga14.intel.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1727757AbgCOCNj (ORCPT ); Sat, 14 Mar 2020 22:13:39 -0400 IronPort-SDR: RJUGGrjMJ89GWXk0Wqjvrh5fOAgfXqGgr7ZEs+2NAlw9PdAzX9DE+nNVPp+pQC1NWU52BAuJuB Poc+azRGR6/Q== X-Amp-Result: SKIPPED(no attachment in message) X-Amp-File-Uploaded: False Received: from orsmga005.jf.intel.com ([10.7.209.41]) by fmsmga103.fm.intel.com with ESMTP/TLS/ECDHE-RSA-AES256-GCM-SHA384; 14 Mar 2020 00:52:25 -0700 IronPort-SDR: 3BP0n+TxPQ1bzWSK5j4bXlycV6foOk8KVYvIpuw29cCDRFK6ruSrrdNywwODy+t6BHzPXXMnb/ LY3W/086UVQA== X-ExtLoop1: 1 X-IronPort-AV: E=Sophos;i="5.70,551,1574150400"; d="scan'208";a="416537645" Received: from lxy-clx-4s.sh.intel.com ([10.239.43.160]) by orsmga005.jf.intel.com with ESMTP; 14 Mar 2020 00:52:21 -0700 From: Xiaoyao Li To: Thomas Gleixner , Ingo Molnar , Borislav Petkov , hpa@zytor.com, Paolo Bonzini , Sean Christopherson , Andy Lutomirski , tony.luck@intel.com Cc: peterz@infradead.org, fenghua.yu@intel.com, Arvind Sankar , Vitaly Kuznetsov , Jim Mattson , x86@kernel.org, kvm@vger.kernel.org, linux-kernel@vger.kernel.org, Xiaoyao Li Subject: [PATCH v4 09/10] kvm: vmx: Enable MSR_TEST_CTRL for intel guest Date: Sat, 14 Mar 2020 15:34:13 +0800 Message-Id: <20200314073414.184213-10-xiaoyao.li@intel.com> X-Mailer: git-send-email 2.20.1 In-Reply-To: <20200314073414.184213-1-xiaoyao.li@intel.com> References: <20200314073414.184213-1-xiaoyao.li@intel.com> MIME-Version: 1.0 Sender: kvm-owner@vger.kernel.org Precedence: bulk List-ID: X-Mailing-List: kvm@vger.kernel.org Only enabling the read and write zero of MSR_TEST_CTRL. This makes MSR_TEST_CTRL always available for intel guest, but guset cannot write any value to it except zero. This matches the truth that most Intel CPUs support MSR_TEST_CTRL, and it also alleviates the effort to handle wrmsr/rdmsr when exposing split lock detect to guest in the following patch. Signed-off-by: Xiaoyao Li --- arch/x86/kvm/vmx/vmx.c | 10 ++++++++++ arch/x86/kvm/vmx/vmx.h | 1 + 2 files changed, 11 insertions(+) diff --git a/arch/x86/kvm/vmx/vmx.c b/arch/x86/kvm/vmx/vmx.c index 3fb132ad489d..107c873b23c2 100644 --- a/arch/x86/kvm/vmx/vmx.c +++ b/arch/x86/kvm/vmx/vmx.c @@ -1831,6 +1831,9 @@ static int vmx_get_msr(struct kvm_vcpu *vcpu, struct msr_data *msr_info) u32 index; switch (msr_info->index) { + case MSR_TEST_CTRL: + msr_info->data = vmx->msr_test_ctrl; + break; #ifdef CONFIG_X86_64 case MSR_FS_BASE: msr_info->data = vmcs_readl(GUEST_FS_BASE); @@ -1984,6 +1987,12 @@ static int vmx_set_msr(struct kvm_vcpu *vcpu, struct msr_data *msr_info) u32 index; switch (msr_index) { + case MSR_TEST_CTRL: + if (data) + return 1; + + vmx->msr_test_ctrl = data; + break; case MSR_EFER: ret = kvm_set_msr_common(vcpu, msr_info); break; @@ -4283,6 +4292,7 @@ static void vmx_vcpu_reset(struct kvm_vcpu *vcpu, bool init_event) vmx->rmode.vm86_active = 0; vmx->spec_ctrl = 0; + vmx->msr_test_ctrl = 0; vmx->msr_ia32_umwait_control = 0; diff --git a/arch/x86/kvm/vmx/vmx.h b/arch/x86/kvm/vmx/vmx.h index e64da06c7009..f679453dcab8 100644 --- a/arch/x86/kvm/vmx/vmx.h +++ b/arch/x86/kvm/vmx/vmx.h @@ -225,6 +225,7 @@ struct vcpu_vmx { #endif u64 spec_ctrl; + u64 msr_test_ctrl; u32 msr_ia32_umwait_control; u32 secondary_exec_control; From patchwork Sat Mar 14 07:34:14 2020 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Xiaoyao Li X-Patchwork-Id: 11438581 Return-Path: Received: from mail.kernel.org (pdx-korg-mail-1.web.codeaurora.org [172.30.200.123]) by pdx-korg-patchwork-2.web.codeaurora.org (Postfix) with ESMTP id 8FA0192C for ; Sun, 15 Mar 2020 02:14:01 +0000 (UTC) Received: from vger.kernel.org (vger.kernel.org [209.132.180.67]) by mail.kernel.org (Postfix) with ESMTP id 713C42077D for ; Sun, 15 Mar 2020 02:14:01 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1727827AbgCOCN6 (ORCPT ); Sat, 14 Mar 2020 22:13:58 -0400 Received: from mga14.intel.com ([192.55.52.115]:41896 "EHLO mga14.intel.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1727768AbgCOCNk (ORCPT ); Sat, 14 Mar 2020 22:13:40 -0400 IronPort-SDR: oJJ/f+jdKWjiRk7IIJZvT5KZnFT3Jl6IfgpRpIpWu+ZdkpcXGWnOItn60oBUGUuX5dKqXzkqf+ a9qnEexyfEdg== X-Amp-Result: SKIPPED(no attachment in message) X-Amp-File-Uploaded: False Received: from orsmga005.jf.intel.com ([10.7.209.41]) by fmsmga103.fm.intel.com with ESMTP/TLS/ECDHE-RSA-AES256-GCM-SHA384; 14 Mar 2020 00:52:29 -0700 IronPort-SDR: 8rOke19cLw6C9Xawh4pfWyou2nS8E9J7f0lqt1yRlMBfjdGJRWiKE4IkQfGDD3j95Cwioylc8q xpP59Rra3jZw== X-ExtLoop1: 1 X-IronPort-AV: E=Sophos;i="5.70,551,1574150400"; d="scan'208";a="416537662" Received: from lxy-clx-4s.sh.intel.com ([10.239.43.160]) by orsmga005.jf.intel.com with ESMTP; 14 Mar 2020 00:52:25 -0700 From: Xiaoyao Li To: Thomas Gleixner , Ingo Molnar , Borislav Petkov , hpa@zytor.com, Paolo Bonzini , Sean Christopherson , Andy Lutomirski , tony.luck@intel.com Cc: peterz@infradead.org, fenghua.yu@intel.com, Arvind Sankar , Vitaly Kuznetsov , Jim Mattson , x86@kernel.org, kvm@vger.kernel.org, linux-kernel@vger.kernel.org, Xiaoyao Li Subject: [PATCH v4 10/10] x86: vmx: virtualize split lock detection Date: Sat, 14 Mar 2020 15:34:14 +0800 Message-Id: <20200314073414.184213-11-xiaoyao.li@intel.com> X-Mailer: git-send-email 2.20.1 In-Reply-To: <20200314073414.184213-1-xiaoyao.li@intel.com> References: <20200314073414.184213-1-xiaoyao.li@intel.com> MIME-Version: 1.0 Sender: kvm-owner@vger.kernel.org Precedence: bulk List-ID: X-Mailing-List: kvm@vger.kernel.org Due to the fact that MSR_TEST_CTRL is per-core scope, i.e., the sibling threads in the same physical CPU core share the same MSR, only advertising feature split lock detection to guest when SMT is disabled or unsupported, for simplicitly. Below summarizing how guest behaves of different host configuration: sld_fatal - hardware MSR_TEST_CTRL.SLD is always on when vcpu is running, even though guest thinks it sets/clears MSR_TEST_CTRL.SLD bit successfully. i.e., SLD is forced on for guest. sld_warn - hardware MSR_TEST_CTRL.SLD is left on until an #AC is intercepted with MSR_TEST_CTRL.SLD=0 in the guest, at which point normal sld_warn rules apply, i.e., clear MSR_TEST_CTRL.SLD bit and set TIF_SLD. If a vCPU associated with the task does VM-Enter with virtual MSR_TEST_CTRL.SLD=1, TIF_SLD is reset, hardware MSR_TEST_CTRL.SLD is re-set, and cycle begins anew. sld_kvm_only - hardware MSR_TEST_CTRL.SLD is set on VM-Entry and cleared onVM-Exit if guest enables SLD, i.e., guest's virtual MSR_TEST_CTRL.SLD is set. sld_disable - guest cannot see feature split lock detection. Signed-off-by: Xiaoyao Li --- arch/x86/include/asm/cpu.h | 2 ++ arch/x86/kernel/cpu/intel.c | 7 ++++++ arch/x86/kvm/vmx/vmx.c | 45 ++++++++++++++++++++++++++++++++----- arch/x86/kvm/x86.c | 17 +++++++++++--- 4 files changed, 63 insertions(+), 8 deletions(-) diff --git a/arch/x86/include/asm/cpu.h b/arch/x86/include/asm/cpu.h index 2e17315b1fed..284be32aaf87 100644 --- a/arch/x86/include/asm/cpu.h +++ b/arch/x86/include/asm/cpu.h @@ -64,6 +64,7 @@ extern void __init cpu_set_core_cap_bits(struct cpuinfo_x86 *c); extern void switch_to_sld(unsigned long tifn); extern bool handle_user_split_lock(unsigned long ip); extern void sld_msr_set(bool on); +extern void sld_turn_back_on(void); #else static inline bool split_lock_detect_on(void) { return false; } static inline bool split_lock_detect_disabled(void) { return true; } @@ -74,5 +75,6 @@ static inline bool handle_user_split_lock(unsigned long ip) return false; } static inline void sld_msr_set(bool on) {} +static inline void sld_turn_back_on(void) {} #endif #endif /* _ASM_X86_CPU_H */ diff --git a/arch/x86/kernel/cpu/intel.c b/arch/x86/kernel/cpu/intel.c index 8bfe8b07e06e..de46e1d3f1c7 100644 --- a/arch/x86/kernel/cpu/intel.c +++ b/arch/x86/kernel/cpu/intel.c @@ -1120,6 +1120,13 @@ void sld_msr_set(bool on) } EXPORT_SYMBOL_GPL(sld_msr_set); +void sld_turn_back_on(void) +{ + __sld_msr_set(true); + clear_tsk_thread_flag(current, TIF_SLD); +} +EXPORT_SYMBOL_GPL(sld_turn_back_on); + /* * This function is called only when switching between tasks with * different split-lock detection modes. It sets the MSR for the diff --git a/arch/x86/kvm/vmx/vmx.c b/arch/x86/kvm/vmx/vmx.c index 107c873b23c2..058dc6c478bd 100644 --- a/arch/x86/kvm/vmx/vmx.c +++ b/arch/x86/kvm/vmx/vmx.c @@ -1819,6 +1819,22 @@ static int vmx_get_msr_feature(struct kvm_msr_entry *msr) } } +static inline u64 vmx_msr_test_ctrl_valid_bits(struct kvm_vcpu *vcpu) +{ + u64 valid_bits = 0; + + /* + * Note: for guest, feature split lock detection can only be enumerated + * through MSR_IA32_CORE_CAPABILITIES bit. + * The FMS enumeration is invalid. + */ + if (vcpu->arch.core_capabilities & + MSR_IA32_CORE_CAPS_SPLIT_LOCK_DETECT) + valid_bits |= MSR_TEST_CTRL_SPLIT_LOCK_DETECT; + + return valid_bits; +} + /* * Reads an msr value (of 'msr_index') into 'pdata'. * Returns 0 on success, non-0 otherwise. @@ -1988,7 +2004,7 @@ static int vmx_set_msr(struct kvm_vcpu *vcpu, struct msr_data *msr_info) switch (msr_index) { case MSR_TEST_CTRL: - if (data) + if (data & ~vmx_msr_test_ctrl_valid_bits(vcpu)) return 1; vmx->msr_test_ctrl = data; @@ -4625,6 +4641,11 @@ static inline bool guest_cpu_alignment_check_enabled(struct kvm_vcpu *vcpu) (kvm_get_rflags(vcpu) & X86_EFLAGS_AC); } +static inline bool guest_cpu_split_lock_detect_on(struct vcpu_vmx *vmx) +{ + return vmx->msr_test_ctrl & MSR_TEST_CTRL_SPLIT_LOCK_DETECT; +} + static int handle_exception_nmi(struct kvm_vcpu *vcpu) { struct vcpu_vmx *vmx = to_vmx(vcpu); @@ -4721,12 +4742,13 @@ static int handle_exception_nmi(struct kvm_vcpu *vcpu) case AC_VECTOR: /* * Reflect #AC to the guest if it's expecting the #AC, i.e. has - * legacy alignment check enabled. Pre-check host split lock - * support to avoid the VMREADs needed to check legacy #AC, - * i.e. reflect the #AC if the only possible source is legacy - * alignment checks. + * legacy alignment check enabled or split lock detect enabled. + * Pre-check host split lock support to avoid further check of + * guest, i.e. reflect the #AC if host doesn't enable split lock + * detection. */ if (!split_lock_detect_on() || + guest_cpu_split_lock_detect_on(vmx) || guest_cpu_alignment_check_enabled(vcpu)) { kvm_queue_exception_e(vcpu, AC_VECTOR, error_code); return 1; @@ -6619,6 +6641,14 @@ static void vmx_vcpu_run(struct kvm_vcpu *vcpu) */ x86_spec_ctrl_set_guest(vmx->spec_ctrl, 0); + if (!split_lock_detect_disabled() && + guest_cpu_split_lock_detect_on(vmx)) { + if (test_thread_flag(TIF_SLD)) + sld_turn_back_on(); + else if (!split_lock_detect_on()) + sld_msr_set(true); + } + /* L1D Flush includes CPU buffer clear to mitigate MDS */ if (static_branch_unlikely(&vmx_l1d_should_flush)) vmx_l1d_flush(vcpu); @@ -6653,6 +6683,11 @@ static void vmx_vcpu_run(struct kvm_vcpu *vcpu) x86_spec_ctrl_restore_host(vmx->spec_ctrl, 0); + if (!split_lock_detect_disabled() && + guest_cpu_split_lock_detect_on(vmx) && + !split_lock_detect_on()) + sld_msr_set(false); + /* All fields are clean at this point */ if (static_branch_unlikely(&enable_evmcs)) current_evmcs->hv_clean_fields |= diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c index 72d4bfea8864..c956aa180253 100644 --- a/arch/x86/kvm/x86.c +++ b/arch/x86/kvm/x86.c @@ -1162,7 +1162,7 @@ static const u32 msrs_to_save_all[] = { #endif MSR_IA32_TSC, MSR_IA32_CR_PAT, MSR_VM_HSAVE_PA, MSR_IA32_FEAT_CTL, MSR_IA32_BNDCFGS, MSR_TSC_AUX, - MSR_IA32_SPEC_CTRL, + MSR_IA32_SPEC_CTRL, MSR_TEST_CTRL, MSR_IA32_RTIT_CTL, MSR_IA32_RTIT_STATUS, MSR_IA32_RTIT_CR3_MATCH, MSR_IA32_RTIT_OUTPUT_BASE, MSR_IA32_RTIT_OUTPUT_MASK, MSR_IA32_RTIT_ADDR0_A, MSR_IA32_RTIT_ADDR0_B, @@ -1344,7 +1344,12 @@ static u64 kvm_get_arch_capabilities(void) static u64 kvm_get_core_capabilities(void) { - return 0; + u64 data = 0; + + if (!split_lock_detect_disabled() && !cpu_smt_possible()) + data |= MSR_IA32_CORE_CAPS_SPLIT_LOCK_DETECT; + + return data; } static int kvm_get_msr_feature(struct kvm_msr_entry *msr) @@ -2729,7 +2734,8 @@ int kvm_set_msr_common(struct kvm_vcpu *vcpu, struct msr_data *msr_info) vcpu->arch.arch_capabilities = data; break; case MSR_IA32_CORE_CAPS: - if (!msr_info->host_initiated) + if (!msr_info->host_initiated || + data & ~kvm_get_core_capabilities()) return 1; vcpu->arch.core_capabilities = data; break; @@ -5276,6 +5282,11 @@ static void kvm_init_msr_list(void) * to the guests in some cases. */ switch (msrs_to_save_all[i]) { + case MSR_TEST_CTRL: + if (!(kvm_get_core_capabilities() & + MSR_IA32_CORE_CAPS_SPLIT_LOCK_DETECT)) + continue; + break; case MSR_IA32_BNDCFGS: if (!kvm_mpx_supported()) continue;