From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from mail-pg0-f69.google.com (mail-pg0-f69.google.com [74.125.83.69]) by kanga.kvack.org (Postfix) with ESMTP id 572566B0271 for ; Thu, 7 Jun 2018 10:40:44 -0400 (EDT) Received: by mail-pg0-f69.google.com with SMTP id q22-v6so3571186pgv.22 for ; Thu, 07 Jun 2018 07:40:44 -0700 (PDT) Received: from mga18.intel.com (mga18.intel.com. [134.134.136.126]) by mx.google.com with ESMTPS id b60-v6si54342625plc.270.2018.06.07.07.40.42 for (version=TLS1_2 cipher=ECDHE-RSA-AES128-GCM-SHA256 bits=128/128); Thu, 07 Jun 2018 07:40:43 -0700 (PDT) From: Yu-cheng Yu Subject: [PATCH 8/9] x86/cet: Handle shadow stack page fault Date: Thu, 7 Jun 2018 07:37:04 -0700 Message-Id: <20180607143705.3531-9-yu-cheng.yu@intel.com> In-Reply-To: <20180607143705.3531-1-yu-cheng.yu@intel.com> References: <20180607143705.3531-1-yu-cheng.yu@intel.com> Sender: owner-linux-mm@kvack.org List-ID: To: linux-kernel@vger.kernel.org, linux-doc@vger.kernel.org, linux-mm@kvack.org, linux-arch@vger.kernel.org, x86@kernel.org, "H. Peter Anvin" , Thomas Gleixner , Ingo Molnar , "H.J. Lu" , Vedvyas Shanbhogue , "Ravi V. Shankar" , Dave Hansen , Andy Lutomirski , Jonathan Corbet , Oleg Nesterov , Arnd Bergmann , Mike Kravetz Cc: Yu-cheng Yu When a task does fork(), its shadow stack must be duplicated for the child. However, the child may not actually use all pages of of the copied shadow stack. This patch implements a flow that is similar to copy-on-write of an anonymous page, but for shadow stack memory. A shadow stack PTE needs to be RO and dirty. We use this dirty bit requirement to effect the copying of shadow stack pages. In copy_one_pte(), we clear the dirty bit from the shadow stack PTE. On the next shadow stack access to the PTE, a page fault occurs. At that time, we then copy/re-use the page and fix the PTE. Signed-off-by: Yu-cheng Yu --- mm/memory.c | 32 +++++++++++++++++++++++++++++--- 1 file changed, 29 insertions(+), 3 deletions(-) diff --git a/mm/memory.c b/mm/memory.c index 01f5464e0fd2..275c7fb3fc96 100644 --- a/mm/memory.c +++ b/mm/memory.c @@ -1022,7 +1022,7 @@ copy_one_pte(struct mm_struct *dst_mm, struct mm_struct *src_mm, * in the parent and the child */ if (is_cow_mapping(vm_flags)) { - ptep_set_wrprotect(src_mm, addr, src_pte); + ptep_set_wrprotect_flush(vma, addr, src_pte); pte = pte_wrprotect(pte); } @@ -2444,7 +2444,13 @@ static inline void wp_page_reuse(struct vm_fault *vmf) flush_cache_page(vma, vmf->address, pte_pfn(vmf->orig_pte)); entry = pte_mkyoung(vmf->orig_pte); - entry = maybe_mkwrite(pte_mkdirty(entry), vma); + + if (is_shstk_mapping(vma->vm_flags)) + entry = pte_mkdirty_shstk(entry); + else + entry = pte_mkdirty(entry); + + entry = maybe_mkwrite(entry, vma); if (ptep_set_access_flags(vma, vmf->address, vmf->pte, entry, 1)) update_mmu_cache(vma, vmf->address, vmf->pte); pte_unmap_unlock(vmf->pte, vmf->ptl); @@ -2517,7 +2523,11 @@ static int wp_page_copy(struct vm_fault *vmf) } flush_cache_page(vma, vmf->address, pte_pfn(vmf->orig_pte)); entry = mk_pte(new_page, vma->vm_page_prot); - entry = maybe_mkwrite(pte_mkdirty(entry), vma); + if (is_shstk_mapping(vma->vm_flags)) + entry = pte_mkdirty_shstk(entry); + else + entry = pte_mkdirty(entry); + entry = maybe_mkwrite(entry, vma); /* * Clear the pte entry and flush it first, before updating the * pte with the new entry. This will avoid a race condition @@ -3192,6 +3202,14 @@ static int do_anonymous_page(struct vm_fault *vmf) mem_cgroup_commit_charge(page, memcg, false, false); lru_cache_add_active_or_unevictable(page, vma); setpte: + /* + * If this is within a shadow stack mapping, mark + * the PTE dirty. We don't use pte_mkdirty(), + * because the PTE must have _PAGE_DIRTY_HW set. + */ + if (is_shstk_mapping(vma->vm_flags)) + entry = pte_mkdirty_shstk(entry); + set_pte_at(vma->vm_mm, vmf->address, vmf->pte, entry); /* No need to invalidate - it was non-present before */ @@ -3974,6 +3992,14 @@ static int handle_pte_fault(struct vm_fault *vmf) entry = vmf->orig_pte; if (unlikely(!pte_same(*vmf->pte, entry))) goto unlock; + + /* + * Shadow stack PTEs are copy-on-access, so do_wp_page() + * handling on them no matter if we have write fault or not. + */ + if (is_shstk_mapping(vmf->vma->vm_flags)) + return do_wp_page(vmf); + if (vmf->flags & FAULT_FLAG_WRITE) { if (!pte_write(entry)) return do_wp_page(vmf); -- 2.15.1