All of lore.kernel.org
 help / color / mirror / Atom feed
From: Dan Williams <dan.j.williams@intel.com>
To: linux-kernel@vger.kernel.org
Cc: linux-arch@vger.kernel.org, Kees Cook <keescook@chromium.org>,
	kernel-hardening@lists.openwall.com, gregkh@linuxfoundation.org,
	x86@kernel.org, Ingo Molnar <mingo@redhat.com>,
	Al Viro <viro@zeniv.linux.org.uk>,
	"H. Peter Anvin" <hpa@zytor.com>,
	tglx@linutronix.de, torvalds@linux-foundation.org,
	akpm@linux-foundation.org, alan@linux.intel.com
Subject: [PATCH v4 06/10] x86, get_user: use pointer masking to limit speculation
Date: Thu, 18 Jan 2018 16:02:15 -0800	[thread overview]
Message-ID: <151632013524.21271.4516878993060956547.stgit@dwillia2-desk3.amr.corp.intel.com> (raw)
In-Reply-To: <151632009605.21271.11304291057104672116.stgit@dwillia2-desk3.amr.corp.intel.com>

Quoting Linus:

    I do think that it would be a good idea to very expressly document
    the fact that it's not that the user access itself is unsafe. I do
    agree that things like "get_user()" want to be protected, but not
    because of any direct bugs or problems with get_user() and friends,
    but simply because get_user() is an excellent source of a pointer
    that is obviously controlled from a potentially attacking user
    space. So it's a prime candidate for then finding _subsequent_
    accesses that can then be used to perturb the cache.

Unlike the '__get_user' case 'get_user' includes the address limit check
near the pointer de-reference. With that locality the speculation can be
mitigated with pointer narrowing rather than a barrier. Where the
narrowing is performed by:

	cmp %limit, %ptr
	sbb %mask, %mask
	and %mask, %ptr

With respect to speculation the value of %ptr is either less than %limit
or NULL.

Co-developed-by: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Cc: Kees Cook <keescook@chromium.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: "H. Peter Anvin" <hpa@zytor.com>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: x86@kernel.org
Signed-off-by: Dan Williams <dan.j.williams@intel.com>
---
 arch/x86/include/asm/smap.h |   17 +++++++++++++++++
 arch/x86/lib/getuser.S      |    5 +++++
 2 files changed, 22 insertions(+)

diff --git a/arch/x86/include/asm/smap.h b/arch/x86/include/asm/smap.h
index db333300bd4b..2b4ad4c6a226 100644
--- a/arch/x86/include/asm/smap.h
+++ b/arch/x86/include/asm/smap.h
@@ -25,6 +25,23 @@
 
 #include <asm/alternative-asm.h>
 
+/*
+ * MASK_NOSPEC - sanitize the value of a user controlled value with
+ * respect to speculation
+ *
+ * In the get_user path once we have determined that the pointer is
+ * below the current address limit sanitize its value with respect to
+ * speculation. In the case when the pointer is above the address limit
+ * this directs the cpu to speculate with a NULL ptr rather than
+ * something targeting kernel memory.
+ *
+ * assumes CF is set from a previous 'cmp TASK_addr_limit, %ptr'
+ */
+.macro MASK_NOSPEC mask val
+	sbb \mask, \mask
+	and \mask, \val
+.endm
+
 #ifdef CONFIG_X86_SMAP
 
 #define ASM_CLAC \
diff --git a/arch/x86/lib/getuser.S b/arch/x86/lib/getuser.S
index c97d935a29e8..07d0e8a28b17 100644
--- a/arch/x86/lib/getuser.S
+++ b/arch/x86/lib/getuser.S
@@ -40,6 +40,7 @@ ENTRY(__get_user_1)
 	mov PER_CPU_VAR(current_task), %_ASM_DX
 	cmp TASK_addr_limit(%_ASM_DX),%_ASM_AX
 	jae bad_get_user
+	MASK_NOSPEC %_ASM_DX, %_ASM_AX
 	ASM_STAC
 1:	movzbl (%_ASM_AX),%edx
 	xor %eax,%eax
@@ -54,6 +55,7 @@ ENTRY(__get_user_2)
 	mov PER_CPU_VAR(current_task), %_ASM_DX
 	cmp TASK_addr_limit(%_ASM_DX),%_ASM_AX
 	jae bad_get_user
+	MASK_NOSPEC %_ASM_DX, %_ASM_AX
 	ASM_STAC
 2:	movzwl -1(%_ASM_AX),%edx
 	xor %eax,%eax
@@ -68,6 +70,7 @@ ENTRY(__get_user_4)
 	mov PER_CPU_VAR(current_task), %_ASM_DX
 	cmp TASK_addr_limit(%_ASM_DX),%_ASM_AX
 	jae bad_get_user
+	MASK_NOSPEC %_ASM_DX, %_ASM_AX
 	ASM_STAC
 3:	movl -3(%_ASM_AX),%edx
 	xor %eax,%eax
@@ -83,6 +86,7 @@ ENTRY(__get_user_8)
 	mov PER_CPU_VAR(current_task), %_ASM_DX
 	cmp TASK_addr_limit(%_ASM_DX),%_ASM_AX
 	jae bad_get_user
+	MASK_NOSPEC %_ASM_DX, %_ASM_AX
 	ASM_STAC
 4:	movq -7(%_ASM_AX),%rdx
 	xor %eax,%eax
@@ -94,6 +98,7 @@ ENTRY(__get_user_8)
 	mov PER_CPU_VAR(current_task), %_ASM_DX
 	cmp TASK_addr_limit(%_ASM_DX),%_ASM_AX
 	jae bad_get_user_8
+	MASK_NOSPEC %_ASM_DX, %_ASM_AX
 	ASM_STAC
 4:	movl -7(%_ASM_AX),%edx
 5:	movl -3(%_ASM_AX),%ecx

WARNING: multiple messages have this Message-ID (diff)
From: Dan Williams <dan.j.williams@intel.com>
To: linux-kernel@vger.kernel.org
Cc: linux-arch@vger.kernel.org, Kees Cook <keescook@chromium.org>,
	kernel-hardening@lists.openwall.com, gregkh@linuxfoundation.org,
	x86@kernel.org, Ingo Molnar <mingo@redhat.com>,
	Al Viro <viro@zeniv.linux.org.uk>,
	"H. Peter Anvin" <hpa@zytor.com>,
	tglx@linutronix.de, torvalds@linux-foundation.org,
	akpm@linux-foundation.org, alan@linux.intel.com
Subject: [kernel-hardening] [PATCH v4 06/10] x86, get_user: use pointer masking to limit speculation
Date: Thu, 18 Jan 2018 16:02:15 -0800	[thread overview]
Message-ID: <151632013524.21271.4516878993060956547.stgit@dwillia2-desk3.amr.corp.intel.com> (raw)
In-Reply-To: <151632009605.21271.11304291057104672116.stgit@dwillia2-desk3.amr.corp.intel.com>

Quoting Linus:

    I do think that it would be a good idea to very expressly document
    the fact that it's not that the user access itself is unsafe. I do
    agree that things like "get_user()" want to be protected, but not
    because of any direct bugs or problems with get_user() and friends,
    but simply because get_user() is an excellent source of a pointer
    that is obviously controlled from a potentially attacking user
    space. So it's a prime candidate for then finding _subsequent_
    accesses that can then be used to perturb the cache.

Unlike the '__get_user' case 'get_user' includes the address limit check
near the pointer de-reference. With that locality the speculation can be
mitigated with pointer narrowing rather than a barrier. Where the
narrowing is performed by:

	cmp %limit, %ptr
	sbb %mask, %mask
	and %mask, %ptr

With respect to speculation the value of %ptr is either less than %limit
or NULL.

Co-developed-by: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Cc: Kees Cook <keescook@chromium.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: "H. Peter Anvin" <hpa@zytor.com>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: x86@kernel.org
Signed-off-by: Dan Williams <dan.j.williams@intel.com>
---
 arch/x86/include/asm/smap.h |   17 +++++++++++++++++
 arch/x86/lib/getuser.S      |    5 +++++
 2 files changed, 22 insertions(+)

diff --git a/arch/x86/include/asm/smap.h b/arch/x86/include/asm/smap.h
index db333300bd4b..2b4ad4c6a226 100644
--- a/arch/x86/include/asm/smap.h
+++ b/arch/x86/include/asm/smap.h
@@ -25,6 +25,23 @@
 
 #include <asm/alternative-asm.h>
 
+/*
+ * MASK_NOSPEC - sanitize the value of a user controlled value with
+ * respect to speculation
+ *
+ * In the get_user path once we have determined that the pointer is
+ * below the current address limit sanitize its value with respect to
+ * speculation. In the case when the pointer is above the address limit
+ * this directs the cpu to speculate with a NULL ptr rather than
+ * something targeting kernel memory.
+ *
+ * assumes CF is set from a previous 'cmp TASK_addr_limit, %ptr'
+ */
+.macro MASK_NOSPEC mask val
+	sbb \mask, \mask
+	and \mask, \val
+.endm
+
 #ifdef CONFIG_X86_SMAP
 
 #define ASM_CLAC \
diff --git a/arch/x86/lib/getuser.S b/arch/x86/lib/getuser.S
index c97d935a29e8..07d0e8a28b17 100644
--- a/arch/x86/lib/getuser.S
+++ b/arch/x86/lib/getuser.S
@@ -40,6 +40,7 @@ ENTRY(__get_user_1)
 	mov PER_CPU_VAR(current_task), %_ASM_DX
 	cmp TASK_addr_limit(%_ASM_DX),%_ASM_AX
 	jae bad_get_user
+	MASK_NOSPEC %_ASM_DX, %_ASM_AX
 	ASM_STAC
 1:	movzbl (%_ASM_AX),%edx
 	xor %eax,%eax
@@ -54,6 +55,7 @@ ENTRY(__get_user_2)
 	mov PER_CPU_VAR(current_task), %_ASM_DX
 	cmp TASK_addr_limit(%_ASM_DX),%_ASM_AX
 	jae bad_get_user
+	MASK_NOSPEC %_ASM_DX, %_ASM_AX
 	ASM_STAC
 2:	movzwl -1(%_ASM_AX),%edx
 	xor %eax,%eax
@@ -68,6 +70,7 @@ ENTRY(__get_user_4)
 	mov PER_CPU_VAR(current_task), %_ASM_DX
 	cmp TASK_addr_limit(%_ASM_DX),%_ASM_AX
 	jae bad_get_user
+	MASK_NOSPEC %_ASM_DX, %_ASM_AX
 	ASM_STAC
 3:	movl -3(%_ASM_AX),%edx
 	xor %eax,%eax
@@ -83,6 +86,7 @@ ENTRY(__get_user_8)
 	mov PER_CPU_VAR(current_task), %_ASM_DX
 	cmp TASK_addr_limit(%_ASM_DX),%_ASM_AX
 	jae bad_get_user
+	MASK_NOSPEC %_ASM_DX, %_ASM_AX
 	ASM_STAC
 4:	movq -7(%_ASM_AX),%rdx
 	xor %eax,%eax
@@ -94,6 +98,7 @@ ENTRY(__get_user_8)
 	mov PER_CPU_VAR(current_task), %_ASM_DX
 	cmp TASK_addr_limit(%_ASM_DX),%_ASM_AX
 	jae bad_get_user_8
+	MASK_NOSPEC %_ASM_DX, %_ASM_AX
 	ASM_STAC
 4:	movl -7(%_ASM_AX),%edx
 5:	movl -3(%_ASM_AX),%ecx

  parent reply	other threads:[~2018-01-19  0:02 UTC|newest]

Thread overview: 70+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2018-01-19  0:01 [PATCH v4 00/10] prevent bounds-check bypass via speculative execution Dan Williams
2018-01-19  0:01 ` [kernel-hardening] " Dan Williams
2018-01-19  0:01 ` Dan Williams
2018-01-19  0:01 ` [PATCH v4 01/10] Documentation: document array_ptr Dan Williams
2018-01-19  0:01   ` [kernel-hardening] " Dan Williams
2018-01-19  0:01 ` [PATCH v4 02/10] asm/nospec, array_ptr: sanitize speculative array de-references Dan Williams
2018-01-19  0:01   ` [kernel-hardening] " Dan Williams
2018-01-19 10:20   ` Jann Horn
2018-01-19 17:48     ` Adam Sampson
2018-01-19 17:48       ` Adam Sampson
2018-01-19 17:48       ` Adam Sampson
2018-01-19 17:48       ` Adam Sampson
2018-01-19 18:12       ` Dan Williams
2018-01-19 18:18         ` Will Deacon
2018-01-19 18:18           ` Will Deacon
2018-01-19 18:26           ` [kernel-hardening] " Dan Williams
2018-01-19 18:18     ` Linus Torvalds
2018-01-19 18:18       ` Linus Torvalds
2018-01-19 20:55       ` [kernel-hardening] " Dan Williams
2018-01-25  7:09   ` Cyril Novikov
2018-01-25  7:09     ` [kernel-hardening] " Cyril Novikov
2018-01-25 22:37     ` Dan Williams
2018-01-25 22:37       ` [kernel-hardening] " Dan Williams
2018-01-19  0:01 ` [PATCH v4 03/10] x86: implement array_ptr_mask() Dan Williams
2018-01-19  0:01   ` [kernel-hardening] " Dan Williams
2018-01-19  0:01 ` [PATCH v4 04/10] x86: introduce __uaccess_begin_nospec and ifence Dan Williams
2018-01-19  0:01   ` [kernel-hardening] " Dan Williams
2018-01-19  0:02 ` [PATCH v4 05/10] x86, __get_user: use __uaccess_begin_nospec Dan Williams
2018-01-19  0:02   ` [kernel-hardening] " Dan Williams
2018-01-19  0:02 ` Dan Williams [this message]
2018-01-19  0:02   ` [kernel-hardening] [PATCH v4 06/10] x86, get_user: use pointer masking to limit speculation Dan Williams
2018-01-19  0:02 ` [PATCH v4 07/10] x86: narrow out of bounds syscalls to sys_read under speculation Dan Williams
2018-01-19  0:02   ` [kernel-hardening] " Dan Williams
2018-01-24 14:40   ` Jiri Slaby
2018-01-24 14:40     ` [kernel-hardening] " Jiri Slaby
2018-02-06 19:29   ` Luis Henriques
2018-02-06 19:48     ` Dan Williams
2018-02-06 20:26       ` Linus Torvalds
2018-02-06 20:37         ` Dan Williams
2018-02-06 20:42           ` Linus Torvalds
2018-02-06 20:43             ` Linus Torvalds
2018-02-06 20:49             ` Andy Lutomirski
2018-02-06 20:58               ` Linus Torvalds
2018-02-06 21:37                 ` Dan Williams
2018-02-06 22:52                   ` Linus Torvalds
2018-02-07  0:33                     ` Dan Williams
2018-02-07  1:23                       ` Linus Torvalds
2018-02-06 22:51       ` Luis Henriques
2018-02-06 22:51         ` Luis Henriques
2018-01-19  0:02 ` [PATCH v4 08/10] vfs, fdtable: prevent bounds-check bypass via speculative execution Dan Williams
2018-01-19  0:02   ` [kernel-hardening] " Dan Williams
2018-01-19  0:02 ` [PATCH v4 09/10] kvm, x86: fix spectre-v1 mitigation Dan Williams
2018-01-19  0:02   ` [kernel-hardening] " Dan Williams
2018-01-19  8:42   ` Paolo Bonzini
2018-01-19  8:42     ` [kernel-hardening] " Paolo Bonzini
2018-01-19  0:02 ` [PATCH v4 10/10] nl80211: sanitize array index in parse_txq_params Dan Williams
2018-01-19  0:02   ` [kernel-hardening] " Dan Williams
2018-01-19  0:02   ` Dan Williams
2018-01-21 10:37   ` Johannes Berg
2018-01-21 10:37     ` [kernel-hardening] " Johannes Berg
2018-01-21 10:37     ` Johannes Berg
2018-01-20  6:58 ` [PATCH v4 00/10] prevent bounds-check bypass via speculative execution Dan Williams
2018-01-20  6:58   ` [kernel-hardening] " Dan Williams
2018-01-20  6:58   ` Dan Williams
2018-01-20 16:56   ` Alexei Starovoitov
2018-01-20 16:56     ` [kernel-hardening] " Alexei Starovoitov
2018-01-20 16:56     ` Alexei Starovoitov
2018-01-20 17:07     ` Alexei Starovoitov
2018-01-20 17:07       ` [kernel-hardening] " Alexei Starovoitov
2018-01-20 17:07       ` Alexei Starovoitov

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=151632013524.21271.4516878993060956547.stgit@dwillia2-desk3.amr.corp.intel.com \
    --to=dan.j.williams@intel.com \
    --cc=akpm@linux-foundation.org \
    --cc=alan@linux.intel.com \
    --cc=gregkh@linuxfoundation.org \
    --cc=hpa@zytor.com \
    --cc=keescook@chromium.org \
    --cc=kernel-hardening@lists.openwall.com \
    --cc=linux-arch@vger.kernel.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=mingo@redhat.com \
    --cc=tglx@linutronix.de \
    --cc=torvalds@linux-foundation.org \
    --cc=viro@zeniv.linux.org.uk \
    --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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.