linux-mm.kvack.org archive mirror
 help / color / mirror / Atom feed
From: Yu-cheng Yu <yu-cheng.yu@intel.com>
To: x86@kernel.org, "H. Peter Anvin" <hpa@zytor.com>,
	Thomas Gleixner <tglx@linutronix.de>,
	Ingo Molnar <mingo@redhat.com>,
	linux-kernel@vger.kernel.org, linux-doc@vger.kernel.org,
	linux-mm@kvack.org, linux-arch@vger.kernel.org,
	linux-api@vger.kernel.org, Arnd Bergmann <arnd@arndb.de>,
	Andy Lutomirski <luto@kernel.org>,
	Balbir Singh <bsingharora@gmail.com>,
	Borislav Petkov <bp@alien8.de>,
	Cyrill Gorcunov <gorcunov@gmail.com>,
	Dave Hansen <dave.hansen@linux.intel.com>,
	Eugene Syromiatnikov <esyr@redhat.com>,
	Florian Weimer <fweimer@redhat.com>,
	"H.J. Lu" <hjl.tools@gmail.com>, Jann Horn <jannh@google.com>,
	Jonathan Corbet <corbet@lwn.net>,
	Kees Cook <keescook@chromium.org>,
	Mike Kravetz <mike.kravetz@oracle.com>,
	Nadav Amit <nadav.amit@gmail.com>,
	Oleg Nesterov <oleg@redhat.com>, Pavel Machek <pavel@ucw.cz>,
	Peter Zijlstra <peterz@infradead.org>,
	Randy Dunlap <rdunlap@infradead.org>,
	"Ravi V. Shankar" <ravi.v.shankar@intel.com>,
	Vedvyas Shanbhogue <vedvyas.shanbhogue@intel.com>,
	Dave Martin <Dave.Martin@arm.com>,
	x86-patch-review@intel.com
Cc: Yu-cheng Yu <yu-cheng.yu@intel.com>
Subject: [RFC PATCH v9 12/27] x86/mm: Modify ptep_set_wrprotect and pmdp_set_wrprotect for _PAGE_DIRTY_SW
Date: Wed,  5 Feb 2020 10:19:20 -0800	[thread overview]
Message-ID: <20200205181935.3712-13-yu-cheng.yu@intel.com> (raw)
In-Reply-To: <20200205181935.3712-1-yu-cheng.yu@intel.com>

When Shadow Stack (SHSTK) is enabled, the [R/O + PAGE_DIRTY_HW] setting is
reserved only for SHSTK.  Non-Shadow Stack R/O PTEs are
[R/O + PAGE_DIRTY_SW].

When a PTE goes from [R/W + PAGE_DIRTY_HW] to [R/O + PAGE_DIRTY_SW], it
could become a transient SHSTK PTE in two cases.

The first case is that some processors can start a write but end up seeing
a read-only PTE by the time they get to the Dirty bit, creating a transient
SHSTK PTE.  However, this will not occur on processors supporting SHSTK
therefore we don't need a TLB flush here.

The second case is that when the software, without atomic, tests & replaces
PAGE_DIRTY_HW with PAGE_DIRTY_SW, a transient SHSTK PTE can exist.  This is
prevented with cmpxchg.

Dave Hansen, Jann Horn, Andy Lutomirski, and Peter Zijlstra provided many
insights to the issue.  Jann Horn provided the cmpxchg solution.

v9:
- Change compile-time conditionals to runtime checks.
- Fix parameters of try_cmpxchg(): change pte_t/pmd_t to
  pte_t.pte/pmd_t.pmd.

v4:
- Implement try_cmpxchg().

Signed-off-by: Yu-cheng Yu <yu-cheng.yu@intel.com>
---
 arch/x86/include/asm/pgtable.h | 66 ++++++++++++++++++++++++++++++++++
 1 file changed, 66 insertions(+)

diff --git a/arch/x86/include/asm/pgtable.h b/arch/x86/include/asm/pgtable.h
index 2733e7ec16b3..43cb27379208 100644
--- a/arch/x86/include/asm/pgtable.h
+++ b/arch/x86/include/asm/pgtable.h
@@ -1253,6 +1253,39 @@ static inline pte_t ptep_get_and_clear_full(struct mm_struct *mm,
 static inline void ptep_set_wrprotect(struct mm_struct *mm,
 				      unsigned long addr, pte_t *ptep)
 {
+	/*
+	 * Some processors can start a write, but end up seeing a read-only
+	 * PTE by the time they get to the Dirty bit.  In this case, they
+	 * will set the Dirty bit, leaving a read-only, Dirty PTE which
+	 * looks like a Shadow Stack PTE.
+	 *
+	 * However, this behavior has been improved and will not occur on
+	 * processors supporting Shadow Stack.  Without this guarantee, a
+	 * transition to a non-present PTE and flush the TLB would be
+	 * needed.
+	 *
+	 * When changing a writable PTE to read-only and if the PTE has
+	 * _PAGE_DIRTY_HW set, we move that bit to _PAGE_DIRTY_SW so that
+	 * the PTE is not a valid Shadow Stack PTE.
+	 */
+#ifdef CONFIG_X86_64
+	if (static_cpu_has(X86_FEATURE_SHSTK)) {
+		pte_t new_pte, pte = READ_ONCE(*ptep);
+
+		do {
+			/*
+			 * This is the same as moving _PAGE_DIRTY_HW
+			 * to _PAGE_DIRTY_SW.
+			 */
+			new_pte = pte_wrprotect(pte);
+			new_pte.pte |= (new_pte.pte & _PAGE_DIRTY_HW) >>
+					_PAGE_BIT_DIRTY_HW << _PAGE_BIT_DIRTY_SW;
+			new_pte.pte &= ~_PAGE_DIRTY_HW;
+		} while (!try_cmpxchg(&ptep->pte, &pte.pte, new_pte.pte));
+
+		return;
+	}
+#endif
 	clear_bit(_PAGE_BIT_RW, (unsigned long *)&ptep->pte);
 }
 
@@ -1303,6 +1336,39 @@ static inline pud_t pudp_huge_get_and_clear(struct mm_struct *mm,
 static inline void pmdp_set_wrprotect(struct mm_struct *mm,
 				      unsigned long addr, pmd_t *pmdp)
 {
+	/*
+	 * Some processors can start a write, but end up seeing a read-only
+	 * PMD by the time they get to the Dirty bit.  In this case, they
+	 * will set the Dirty bit, leaving a read-only, Dirty PMD which
+	 * looks like a Shadow Stack PMD.
+	 *
+	 * However, this behavior has been improved and will not occur on
+	 * processors supporting Shadow Stack.  Without this guarantee, a
+	 * transition to a non-present PMD and flush the TLB would be
+	 * needed.
+	 *
+	 * When changing a writable PMD to read-only and if the PMD has
+	 * _PAGE_DIRTY_HW set, we move that bit to _PAGE_DIRTY_SW so that
+	 * the PMD is not a valid Shadow Stack PMD.
+	 */
+#ifdef CONFIG_X86_64
+	if (static_cpu_has(X86_FEATURE_SHSTK)) {
+		pmd_t new_pmd, pmd = READ_ONCE(*pmdp);
+
+		do {
+			/*
+			 * This is the same as moving _PAGE_DIRTY_HW
+			 * to _PAGE_DIRTY_SW.
+			 */
+			new_pmd = pmd_wrprotect(pmd);
+			new_pmd.pmd |= (new_pmd.pmd & _PAGE_DIRTY_HW) >>
+					_PAGE_BIT_DIRTY_HW << _PAGE_BIT_DIRTY_SW;
+			new_pmd.pmd &= ~_PAGE_DIRTY_HW;
+		} while (!try_cmpxchg(&pmdp->pmd, &pmd.pmd, new_pmd.pmd));
+
+		return;
+	}
+#endif
 	clear_bit(_PAGE_BIT_RW, (unsigned long *)pmdp);
 }
 
-- 
2.21.0



  parent reply	other threads:[~2020-02-05 18:21 UTC|newest]

Thread overview: 107+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2020-02-05 18:19 [RFC PATCH v9 00/27] Control-flow Enforcement: Shadow Stack Yu-cheng Yu
2020-02-05 18:19 ` [RFC PATCH v9 01/27] Documentation/x86: Add CET description Yu-cheng Yu
2020-02-06  0:16   ` Randy Dunlap
2020-02-06 20:17     ` Yu-cheng Yu
2020-02-25 20:02   ` Kees Cook
2020-02-28 15:55     ` Yu-cheng Yu
2020-02-26 17:57   ` Dave Hansen
2020-03-09 17:00     ` Yu-cheng Yu
2020-03-09 17:21       ` Dave Hansen
2020-03-09 19:27         ` Yu-cheng Yu
2020-03-09 19:35           ` Dave Hansen
2020-03-09 19:50             ` H.J. Lu
2020-03-09 20:16               ` Andy Lutomirski
2020-03-09 20:54                 ` H.J. Lu
2020-03-09 20:59                   ` Dave Hansen
2020-03-09 21:12                     ` H.J. Lu
2020-03-09 22:02                       ` Andy Lutomirski
2020-03-09 22:19                       ` Dave Hansen
2020-03-09 23:11                         ` H.J. Lu
2020-03-09 23:20                           ` Dave Hansen
2020-03-09 23:51                             ` H.J. Lu
2020-03-09 23:59                               ` Andy Lutomirski
2020-03-10  0:08                                 ` H.J. Lu
2020-03-10  1:21                                   ` Andy Lutomirski
2020-03-10  2:13                                     ` H.J. Lu
2020-02-05 18:19 ` [RFC PATCH v9 02/27] x86/cpufeatures: Add CET CPU feature flags for Control-flow Enforcement Technology (CET) Yu-cheng Yu
2020-02-25 20:02   ` Kees Cook
2020-02-05 18:19 ` [RFC PATCH v9 03/27] x86/fpu/xstate: Introduce CET MSR XSAVES supervisor states Yu-cheng Yu
2020-02-25 20:04   ` Kees Cook
2020-02-05 18:19 ` [RFC PATCH v9 04/27] x86/cet: Add control-protection fault handler Yu-cheng Yu
2020-02-25 20:06   ` Kees Cook
2020-02-26 17:10   ` Dave Hansen
2020-03-05 20:44     ` Yu-cheng Yu
2020-02-05 18:19 ` [RFC PATCH v9 05/27] x86/cet/shstk: Add Kconfig option for user-mode Shadow Stack protection Yu-cheng Yu
2020-02-25 20:07   ` Kees Cook
2020-02-26 17:03   ` Dave Hansen
2020-02-26 19:57     ` Pavel Machek
2020-03-05 20:38     ` Yu-cheng Yu
2020-02-26 18:05   ` Dave Hansen
2020-02-27  1:02     ` H.J. Lu
2020-02-27  1:16       ` Dave Hansen
2020-02-27  2:11         ` H.J. Lu
2020-02-27  3:57           ` Andy Lutomirski
2020-02-27 18:03             ` Dave Hansen
2020-03-06 18:37     ` Yu-cheng Yu
2020-03-06 19:02       ` Dave Hansen
2020-03-06 21:16         ` Yu-cheng Yu
2020-02-05 18:19 ` [RFC PATCH v9 06/27] mm: Introduce VM_SHSTK for Shadow Stack memory Yu-cheng Yu
2020-02-25 20:07   ` Kees Cook
2020-02-26 18:07   ` Dave Hansen
2020-02-05 18:19 ` [RFC PATCH v9 07/27] Add guard pages around a Shadow Stack Yu-cheng Yu
2020-02-25 20:11   ` Kees Cook
2020-02-26 18:17   ` Dave Hansen
2020-02-05 18:19 ` [RFC PATCH v9 08/27] x86/mm: Change _PAGE_DIRTY to _PAGE_DIRTY_HW Yu-cheng Yu
2020-02-25 20:12   ` Kees Cook
2020-02-26 18:20   ` Dave Hansen
2020-02-05 18:19 ` [RFC PATCH v9 09/27] x86/mm: Introduce _PAGE_DIRTY_SW Yu-cheng Yu
2020-02-25 20:12   ` Kees Cook
2020-02-26 21:35   ` Dave Hansen
2020-04-01 19:08     ` Yu-cheng Yu
2020-04-01 19:22       ` Dave Hansen
2020-02-05 18:19 ` [RFC PATCH v9 10/27] x86/mm: Update pte_modify, pmd_modify, and _PAGE_CHG_MASK for _PAGE_DIRTY_SW Yu-cheng Yu
2020-02-26 22:02   ` Dave Hansen
2020-02-05 18:19 ` [RFC PATCH v9 11/27] drm/i915/gvt: Change _PAGE_DIRTY to _PAGE_DIRTY_BITS Yu-cheng Yu
2020-02-25 20:13   ` Kees Cook
2020-02-26 22:04   ` Dave Hansen
2020-04-03 15:42     ` Yu-cheng Yu
2020-02-05 18:19 ` Yu-cheng Yu [this message]
2020-02-25 20:14   ` [RFC PATCH v9 12/27] x86/mm: Modify ptep_set_wrprotect and pmdp_set_wrprotect for _PAGE_DIRTY_SW Kees Cook
2020-02-26 22:20   ` Dave Hansen
2020-02-05 18:19 ` [RFC PATCH v9 13/27] x86/mm: Shadow Stack page fault error checking Yu-cheng Yu
2020-02-25 20:16   ` Kees Cook
2020-02-26 22:47   ` Dave Hansen
2020-02-05 18:19 ` [RFC PATCH v9 14/27] mm: Handle Shadow Stack page fault Yu-cheng Yu
2020-02-25 20:20   ` Kees Cook
2020-03-05 18:30     ` Yu-cheng Yu
2020-02-27  0:08   ` Dave Hansen
2020-04-07 18:14     ` Yu-cheng Yu
2020-04-07 22:21       ` Dave Hansen
2020-04-08 18:18         ` Yu-cheng Yu
2020-02-05 18:19 ` [RFC PATCH v9 15/27] mm: Handle THP/HugeTLB " Yu-cheng Yu
2020-02-25 20:59   ` Kees Cook
2020-03-13 22:00     ` Yu-cheng Yu
2020-02-05 18:19 ` [RFC PATCH v9 16/27] mm: Update can_follow_write_pte() for Shadow Stack Yu-cheng Yu
2020-02-27  0:34   ` Dave Hansen
2020-02-05 18:19 ` [RFC PATCH v9 17/27] x86/cet/shstk: User-mode Shadow Stack support Yu-cheng Yu
2020-02-25 21:07   ` Kees Cook
2020-02-27  0:55   ` Dave Hansen
2020-02-05 18:19 ` [RFC PATCH v9 18/27] x86/cet/shstk: Introduce WRUSS instruction Yu-cheng Yu
2020-02-25 21:10   ` Kees Cook
2020-03-05 18:39     ` Yu-cheng Yu
2020-02-05 18:19 ` [RFC PATCH v9 19/27] x86/cet/shstk: Handle signals for Shadow Stack Yu-cheng Yu
2020-02-25 21:17   ` Kees Cook
2020-02-05 18:19 ` [RFC PATCH v9 20/27] ELF: UAPI and Kconfig additions for ELF program properties Yu-cheng Yu
2020-02-05 18:19 ` [RFC PATCH v9 21/27] binfmt_elf: Define GNU_PROPERTY_X86_FEATURE_1_AND Yu-cheng Yu
2020-02-25 21:18   ` Kees Cook
2020-02-05 18:19 ` [RFC PATCH v9 22/27] ELF: Add ELF program property parsing support Yu-cheng Yu
2020-02-25 21:20   ` Kees Cook
2020-02-05 18:19 ` [RFC PATCH v9 23/27] ELF: Introduce arch_setup_elf_property() Yu-cheng Yu
2020-02-05 18:19 ` [RFC PATCH v9 24/27] x86/cet/shstk: ELF header parsing for Shadow Stack Yu-cheng Yu
2020-02-25 21:22   ` Kees Cook
2020-02-05 18:19 ` [RFC PATCH v9 25/27] x86/cet/shstk: Handle thread " Yu-cheng Yu
2020-02-25 21:29   ` Kees Cook
2020-03-25 21:51     ` Yu-cheng Yu
2020-02-05 18:19 ` [RFC PATCH v9 26/27] mm/mmap: Add Shadow Stack pages to memory accounting Yu-cheng Yu
2020-02-05 18:19 ` [RFC PATCH v9 27/27] x86/cet/shstk: Add arch_prctl functions for Shadow Stack Yu-cheng Yu
2020-02-25 21:31 ` [RFC PATCH v9 00/27] Control-flow Enforcement: " Kees Cook

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20200205181935.3712-13-yu-cheng.yu@intel.com \
    --to=yu-cheng.yu@intel.com \
    --cc=Dave.Martin@arm.com \
    --cc=arnd@arndb.de \
    --cc=bp@alien8.de \
    --cc=bsingharora@gmail.com \
    --cc=corbet@lwn.net \
    --cc=dave.hansen@linux.intel.com \
    --cc=esyr@redhat.com \
    --cc=fweimer@redhat.com \
    --cc=gorcunov@gmail.com \
    --cc=hjl.tools@gmail.com \
    --cc=hpa@zytor.com \
    --cc=jannh@google.com \
    --cc=keescook@chromium.org \
    --cc=linux-api@vger.kernel.org \
    --cc=linux-arch@vger.kernel.org \
    --cc=linux-doc@vger.kernel.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-mm@kvack.org \
    --cc=luto@kernel.org \
    --cc=mike.kravetz@oracle.com \
    --cc=mingo@redhat.com \
    --cc=nadav.amit@gmail.com \
    --cc=oleg@redhat.com \
    --cc=pavel@ucw.cz \
    --cc=peterz@infradead.org \
    --cc=ravi.v.shankar@intel.com \
    --cc=rdunlap@infradead.org \
    --cc=tglx@linutronix.de \
    --cc=vedvyas.shanbhogue@intel.com \
    --cc=x86-patch-review@intel.com \
    --cc=x86@kernel.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).