From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org X-Spam-Level: X-Spam-Status: No, score=-0.8 required=3.0 tests=HEADER_FROM_DIFFERENT_DOMAINS, MAILING_LIST_MULTI,SPF_PASS autolearn=ham autolearn_force=no version=3.4.0 Received: from mail.kernel.org (mail.kernel.org [198.145.29.99]) by smtp.lore.kernel.org (Postfix) with ESMTP id 2093EC433F4 for ; Fri, 31 Aug 2018 16:24:37 +0000 (UTC) Received: from vger.kernel.org (vger.kernel.org [209.132.180.67]) by mail.kernel.org (Postfix) with ESMTP id D760A2077C for ; Fri, 31 Aug 2018 16:24:36 +0000 (UTC) DMARC-Filter: OpenDMARC Filter v1.3.2 mail.kernel.org D760A2077C Authentication-Results: mail.kernel.org; dmarc=fail (p=none dis=none) header.from=intel.com Authentication-Results: mail.kernel.org; spf=none smtp.mailfrom=linux-kernel-owner@vger.kernel.org Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1728174AbeHaUcu (ORCPT ); Fri, 31 Aug 2018 16:32:50 -0400 Received: from mga04.intel.com ([192.55.52.120]:34744 "EHLO mga04.intel.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1727268AbeHaUcu (ORCPT ); Fri, 31 Aug 2018 16:32:50 -0400 X-Amp-Result: SKIPPED(no attachment in message) X-Amp-File-Uploaded: False Received: from orsmga004.jf.intel.com ([10.7.209.38]) by fmsmga104.fm.intel.com with ESMTP/TLS/DHE-RSA-AES256-GCM-SHA384; 31 Aug 2018 09:24:35 -0700 X-ExtLoop1: 1 X-IronPort-AV: E=Sophos;i="5.53,312,1531810800"; d="scan'208";a="229243768" Received: from 2b52.sc.intel.com ([143.183.136.52]) by orsmga004.jf.intel.com with ESMTP; 31 Aug 2018 09:24:34 -0700 Message-ID: <1535732418.3789.7.camel@intel.com> Subject: Re: [RFC PATCH v3 06/24] x86/cet: Control protection exception handler From: Yu-cheng Yu To: Jann Horn Cc: the arch/x86 maintainers , "H . Peter Anvin" , Thomas Gleixner , Ingo Molnar , kernel list , linux-doc@vger.kernel.org, Linux-MM , linux-arch , Linux API , Arnd Bergmann , Andy Lutomirski , Balbir Singh , Cyrill Gorcunov , Dave Hansen , Florian Weimer , hjl.tools@gmail.com, Jonathan Corbet , keescook@chromium.org, Mike Kravetz , Nadav Amit , Oleg Nesterov , Pavel Machek , Peter Zijlstra , ravi.v.shankar@intel.com, vedvyas.shanbhogue@intel.com Date: Fri, 31 Aug 2018 09:20:18 -0700 In-Reply-To: References: <20180830143904.3168-1-yu-cheng.yu@intel.com> <20180830143904.3168-7-yu-cheng.yu@intel.com> Content-Type: text/plain; charset="UTF-8" X-Mailer: Evolution 3.18.5.2-0ubuntu3.2 Mime-Version: 1.0 Content-Transfer-Encoding: 8bit Sender: linux-kernel-owner@vger.kernel.org Precedence: bulk List-ID: X-Mailing-List: linux-kernel@vger.kernel.org On Fri, 2018-08-31 at 17:01 +0200, Jann Horn wrote: > Is there a reason why all the code in this patch isn't #ifdef'ed > away > on builds that don't support CET? It looks like the CET handler is > hooked up to the IDT conditionally, but the handler code is always > built? Yes, in idt.c, it should have been: #ifdef CONFIG_X86_64 INTG(X86_TRAP_CP, control_protection), #endif I will fix it. > > +dotraplinkage void > > +do_control_protection(struct pt_regs *regs, long error_code) > > +{ > > +       struct task_struct *tsk; > > + > > +       RCU_LOCKDEP_WARN(!rcu_is_watching(), "entry code didn't > > wake RCU"); > > +       if (notify_die(DIE_TRAP, "control protection fault", regs, > > +                      error_code, X86_TRAP_CP, SIGSEGV) == > > NOTIFY_STOP) > > +               return; > > +       cond_local_irq_enable(regs); > > + > > +       if (!user_mode(regs)) > > +               die("kernel control protection fault", regs, > > error_code); > > + > > +       if (!static_cpu_has(X86_FEATURE_SHSTK) && > > +           !static_cpu_has(X86_FEATURE_IBT)) > > +               WARN_ONCE(1, "CET is disabled but got control " > > +                         "protection fault\n"); > > + > > +       tsk = current; > > +       tsk->thread.error_code = error_code; > > +       tsk->thread.trap_nr = X86_TRAP_CP; > > + > > +       if (show_unhandled_signals && unhandled_signal(tsk, > > SIGSEGV) && > > +           printk_ratelimit()) { > > +               unsigned int max_err; > > + > > +               max_err = ARRAY_SIZE(control_protection_err) - 1; > > +               if ((error_code < 0) || (error_code > max_err)) > > +                       error_code = 0; > > +               pr_info("%s[%d] control protection ip:%lx sp:%lx > > error:%lx(%s)", > > +                       tsk->comm, task_pid_nr(tsk), > > +                       regs->ip, regs->sp, error_code, > > +                       control_protection_err[error_code]); > > +               print_vma_addr(" in ", regs->ip); > Shouldn't this be using KERN_CONT, like other callers of > print_vma_addr(), to get the desired output? I will change it. From mboxrd@z Thu Jan 1 00:00:00 1970 From: Yu-cheng Yu Subject: Re: [RFC PATCH v3 06/24] x86/cet: Control protection exception handler Date: Fri, 31 Aug 2018 09:20:18 -0700 Message-ID: <1535732418.3789.7.camel@intel.com> References: <20180830143904.3168-1-yu-cheng.yu@intel.com> <20180830143904.3168-7-yu-cheng.yu@intel.com> Mime-Version: 1.0 Content-Type: text/plain; charset="UTF-8" Content-Transfer-Encoding: 8bit Return-path: In-Reply-To: Sender: linux-kernel-owner@vger.kernel.org To: Jann Horn Cc: the arch/x86 maintainers , "H . Peter Anvin" , Thomas Gleixner , Ingo Molnar , kernel list , linux-doc@vger.kernel.org, Linux-MM , linux-arch , Linux API , Arnd Bergmann , Andy Lutomirski , Balbir Singh , Cyrill Gorcunov , Dave Hansen , Florian Weimer , hjl.tools@gmail.com, Jonathan Corbet , keescook@chromium.org, Mike Kravetz , Nadav Amit , Oleg Nesterov , Pavel Machek List-Id: linux-api@vger.kernel.org On Fri, 2018-08-31 at 17:01 +0200, Jann Horn wrote: > Is there a reason why all the code in this patch isn't #ifdef'ed > away > on builds that don't support CET? It looks like the CET handler is > hooked up to the IDT conditionally, but the handler code is always > built? Yes, in idt.c, it should have been: #ifdef CONFIG_X86_64 INTG(X86_TRAP_CP, control_protection), #endif I will fix it. > > +dotraplinkage void > > +do_control_protection(struct pt_regs *regs, long error_code) > > +{ > > +       struct task_struct *tsk; > > + > > +       RCU_LOCKDEP_WARN(!rcu_is_watching(), "entry code didn't > > wake RCU"); > > +       if (notify_die(DIE_TRAP, "control protection fault", regs, > > +                      error_code, X86_TRAP_CP, SIGSEGV) == > > NOTIFY_STOP) > > +               return; > > +       cond_local_irq_enable(regs); > > + > > +       if (!user_mode(regs)) > > +               die("kernel control protection fault", regs, > > error_code); > > + > > +       if (!static_cpu_has(X86_FEATURE_SHSTK) && > > +           !static_cpu_has(X86_FEATURE_IBT)) > > +               WARN_ONCE(1, "CET is disabled but got control " > > +                         "protection fault\n"); > > + > > +       tsk = current; > > +       tsk->thread.error_code = error_code; > > +       tsk->thread.trap_nr = X86_TRAP_CP; > > + > > +       if (show_unhandled_signals && unhandled_signal(tsk, > > SIGSEGV) && > > +           printk_ratelimit()) { > > +               unsigned int max_err; > > + > > +               max_err = ARRAY_SIZE(control_protection_err) - 1; > > +               if ((error_code < 0) || (error_code > max_err)) > > +                       error_code = 0; > > +               pr_info("%s[%d] control protection ip:%lx sp:%lx > > error:%lx(%s)", > > +                       tsk->comm, task_pid_nr(tsk), > > +                       regs->ip, regs->sp, error_code, > > +                       control_protection_err[error_code]); > > +               print_vma_addr(" in ", regs->ip); > Shouldn't this be using KERN_CONT, like other callers of > print_vma_addr(), to get the desired output? I will change it. From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from mail-pl1-f197.google.com (mail-pl1-f197.google.com [209.85.214.197]) by kanga.kvack.org (Postfix) with ESMTP id 9C2496B57DB for ; Fri, 31 Aug 2018 12:24:36 -0400 (EDT) Received: by mail-pl1-f197.google.com with SMTP id w11-v6so6261369plq.8 for ; Fri, 31 Aug 2018 09:24:36 -0700 (PDT) Received: from mga14.intel.com (mga14.intel.com. [192.55.52.115]) by mx.google.com with ESMTPS id o12-v6si8977548pls.94.2018.08.31.09.24.35 for (version=TLS1_2 cipher=ECDHE-RSA-AES128-GCM-SHA256 bits=128/128); Fri, 31 Aug 2018 09:24:35 -0700 (PDT) Message-ID: <1535732418.3789.7.camel@intel.com> Subject: Re: [RFC PATCH v3 06/24] x86/cet: Control protection exception handler From: Yu-cheng Yu Date: Fri, 31 Aug 2018 09:20:18 -0700 In-Reply-To: References: <20180830143904.3168-1-yu-cheng.yu@intel.com> <20180830143904.3168-7-yu-cheng.yu@intel.com> Content-Type: text/plain; charset="UTF-8" Mime-Version: 1.0 Content-Transfer-Encoding: 8bit Sender: owner-linux-mm@kvack.org List-ID: To: Jann Horn Cc: the arch/x86 maintainers , "H . Peter Anvin" , Thomas Gleixner , Ingo Molnar , kernel list , linux-doc@vger.kernel.org, Linux-MM , linux-arch , Linux API , Arnd Bergmann , Andy Lutomirski , Balbir Singh , Cyrill Gorcunov , Dave Hansen , Florian Weimer , hjl.tools@gmail.com, Jonathan Corbet , keescook@chromium.org, Mike Kravetz , Nadav Amit , Oleg Nesterov , Pavel Machek , Peter Zijlstra , ravi.v.shankar@intel.com, vedvyas.shanbhogue@intel.com On Fri, 2018-08-31 at 17:01 +0200, Jann Horn wrote: > Is there a reason why all the code in this patch isn't #ifdef'ed > away > on builds that don't support CET? It looks like the CET handler is > hooked up to the IDT conditionally, but the handler code is always > built? Yes, in idt.c, it should have been: #ifdef CONFIG_X86_64 INTG(X86_TRAP_CP, control_protection), #endif I will fix it. > > +dotraplinkage void > > +do_control_protection(struct pt_regs *regs, long error_code) > > +{ > > +A A A A A A A struct task_struct *tsk; > > + > > +A A A A A A A RCU_LOCKDEP_WARN(!rcu_is_watching(), "entry code didn't > > wake RCU"); > > +A A A A A A A if (notify_die(DIE_TRAP, "control protection fault", regs, > > +A A A A A A A A A A A A A A A A A A A A A A error_code, X86_TRAP_CP, SIGSEGV) == > > NOTIFY_STOP) > > +A A A A A A A A A A A A A A A return; > > +A A A A A A A cond_local_irq_enable(regs); > > + > > +A A A A A A A if (!user_mode(regs)) > > +A A A A A A A A A A A A A A A die("kernel control protection fault", regs, > > error_code); > > + > > +A A A A A A A if (!static_cpu_has(X86_FEATURE_SHSTK) && > > +A A A A A A A A A A A !static_cpu_has(X86_FEATURE_IBT)) > > +A A A A A A A A A A A A A A A WARN_ONCE(1, "CET is disabled but got control " > > +A A A A A A A A A A A A A A A A A A A A A A A A A "protection fault\n"); > > + > > +A A A A A A A tsk = current; > > +A A A A A A A tsk->thread.error_code = error_code; > > +A A A A A A A tsk->thread.trap_nr = X86_TRAP_CP; > > + > > +A A A A A A A if (show_unhandled_signals && unhandled_signal(tsk, > > SIGSEGV) && > > +A A A A A A A A A A A printk_ratelimit()) { > > +A A A A A A A A A A A A A A A unsigned int max_err; > > + > > +A A A A A A A A A A A A A A A max_err = ARRAY_SIZE(control_protection_err) - 1; > > +A A A A A A A A A A A A A A A if ((error_code < 0) || (error_code > max_err)) > > +A A A A A A A A A A A A A A A A A A A A A A A error_code = 0; > > +A A A A A A A A A A A A A A A pr_info("%s[%d] control protection ip:%lx sp:%lx > > error:%lx(%s)", > > +A A A A A A A A A A A A A A A A A A A A A A A tsk->comm, task_pid_nr(tsk), > > +A A A A A A A A A A A A A A A A A A A A A A A regs->ip, regs->sp, error_code, > > +A A A A A A A A A A A A A A A A A A A A A A A control_protection_err[error_code]); > > +A A A A A A A A A A A A A A A print_vma_addr(" in ", regs->ip); > Shouldn't this be using KERN_CONT, like other callers of > print_vma_addr(), to get the desired output? I will change it.