Skip to contentSkip to conversation

x86: clear XD_DISABLED flag on Intel to regain NX

PATCH In review Kees Cook 3 participants 8 messages Nov 10, 2010 – Dec 22, 2010 1 message not in the archive

Subsystems (2)
lore
Share & tools
lore

Kees Cook

author
#0 lore raw

v5 cover x86: clear XD_DISABLED flag on Intel to regain NX

Resending with Alan Cox's Acks, at Ingo's request.

Intel CPUs have an additional MSR bit to indicate if the BIOS was configured to disable NX. This bit was traditionally used for operating systems that did not understand how to handle the NX bit. Since Linux understands this, this BIOS flag should be ignored by default.

In a review[1] of reported hardware being used by Ubuntu bug reporters, almost 10% of systems had an incorrectly configured BIOS, leaving their systems unable to use the NX features of their CPU.

This change will clear the MSR_IA32_MISC_ENABLE_XD_DISABLE bit so that NX cannot be inappropriately controlled by the BIOS on Intel CPUs. If, under very strange hardware configurations, NX actually needs to be disabled, "noexec=off" can be used to restore the prior behavior.

-Kees

[1] http://www.outflux.net/blog/archives/2010/02/18/data-mining-for-nx-bit/

---
Changelog:
v2 - rearranged use of verify_cpu()
v3 - resent, show stats on systems that need it
v4 - expanded commit messages, added acks
v5 - added acks

author
#1 lore raw

PATCH v5 1/4 x86: rename verify_cpu_64.S to verify_cpu.S

The code is 32bit already, and can be used in 32bit routines.

---
 arch/x86/boot/compressed/head_64.S                |    2 +-
 arch/x86/kernel/trampoline_64.S                   |    2 +-
 arch/x86/kernel/{verify_cpu_64.S => verify_cpu.S} |    0
 3 files changed, 2 insertions(+), 2 deletions(-)
 rename arch/x86/kernel/{verify_cpu_64.S => verify_cpu.S} (100%)
arch/x86/boot/compressed/head_64.S +1 -1
index 52f85a1..35af09d 100644
@@ -182,7 +182,7 @@ no_longmode:
 	hlt
 	jmp     1b
 
-#include "../../kernel/verify_cpu_64.S"
+#include "../../kernel/verify_cpu.S"
 
 	/*
 	 * Be careful here startup_64 needs to be at a predictable
arch/x86/kernel/trampoline_64.S +1 -1
index 3af2dff..075d130 100644
@@ -127,7 +127,7 @@ startup_64:
 no_longmode:
 	hlt
 	jmp no_longmode
-#include "verify_cpu_64.S"
+#include "verify_cpu.S"
 
 	# Careful these need to be in the same 64K segment as the above;
 tidt:
arch/x86/kernel/verify_cpu.S renamed from arch/x86/kernel/verify_cpu_64.S +0 -0
similarity index 100%
rename from arch/x86/kernel/verify_cpu_64.S
rename to arch/x86/kernel/verify_cpu.S
Signature · 1 line
1.7.2.3
author
#2 lore raw

PATCH v5 2/4 x86: clear XD_DISABLED flag on Intel to regain NX

In reply to Kees Cook

Intel CPUs have an additional MSR bit to indicate if the BIOS was configured to disable the NX cpu feature. This bit was traditionally used for operating systems that did not understand how to handle the NX bit. Since Linux understands this, this BIOS flag should be ignored by default.

In a review[1] of reported hardware being used by Ubuntu bug reporters, almost 10% of systems had an incorrectly configured BIOS, leaving their systems unable to use the NX features of their CPU.

This change will clear the MSR_IA32_MISC_ENABLE_XD_DISABLE bit so that NX cannot be inappropriately controlled by the BIOS on Intel CPUs. If, under very strange hardware configurations, NX actually needs to be disabled, "noexec=off" can be used to restore the prior behavior.

[1] http://www.outflux.net/blog/archives/2010/02/18/data-mining-for-nx-bit/

---
 arch/x86/kernel/verify_cpu.S |   48 +++++++++++++++++++++++++++++++++++-------
 1 files changed, 40 insertions(+), 8 deletions(-)
arch/x86/kernel/verify_cpu.S +40 -8
index 56a8c2a..ccb4136 100644
@@ -7,6 +7,7 @@
  *	Copyright (c) 2007  Andi Kleen ([email protected])
  *	Copyright (c) 2007  Eric Biederman ([email protected])
  *	Copyright (c) 2007  Vivek Goyal ([email protected])
+ *	Copyright (c) 2010  Kees Cook ([email protected])
  *
  * 	This source code is licensed under the GNU General Public License,
  * 	Version 2.  See the file COPYING for more details.
@@ -14,18 +15,16 @@
  *	This is a common code for verification whether CPU supports
  * 	long mode and SSE or not. It is not called directly instead this
  *	file is included at various places and compiled in that context.
- * 	Following are the current usage.
+ *	This file is expected to run in 32bit code.  Currently:
  *
- * 	This file is included by both 16bit and 32bit code.
+ *	arch/x86_64/boot/compressed/head_64.S: Boot cpu verification
+ *	arch/x86_64/kernel/trampoline_64.S: secondary processor verfication
  *
- *	arch/x86_64/boot/setup.S : Boot cpu verification (16bit)
- *	arch/x86_64/boot/compressed/head.S: Boot cpu verification (32bit)
- *	arch/x86_64/kernel/trampoline.S: secondary processor verfication (16bit)
- *	arch/x86_64/kernel/acpi/wakeup.S:Verfication at resume (16bit)
- *
- *	verify_cpu, returns the status of cpu check in register %eax.
+ *	verify_cpu, returns the status of longmode and SSE in register %eax.
  *		0: Success    1: Failure
  *
+ *	On Intel, the XD_DISABLE flag will be cleared as a side-effect.
+ *
  * 	The caller needs to check for the error code and take the action
  * 	appropriately. Either display a message or halt.
  */
@@ -62,8 +61,41 @@ verify_cpu:
 	cmpl	$0x444d4163,%ecx
 	jnz	verify_cpu_noamd
 	mov	$1,%di			# cpu is from AMD
+	jmp	verify_cpu_check
 
 verify_cpu_noamd:
+	cmpl	$0x756e6547,%ebx        # GenuineIntel?
+	jnz	verify_cpu_check
+	cmpl	$0x49656e69,%edx
+	jnz	verify_cpu_check
+	cmpl	$0x6c65746e,%ecx
+	jnz	verify_cpu_check
+
+	# only call IA32_MISC_ENABLE when:
+	# family > 6 || (family == 6 && model >= 0xd)
+	movl	$0x1, %eax		# check CPU family and model
+	cpuid
+	movl	%eax, %ecx
+
+	andl	$0x0ff00f00, %eax	# mask family and extended family
+	shrl	$8, %eax
+	cmpl	$6, %eax
+	ja	verify_cpu_clear_xd	# family > 6, ok
+	jb	verify_cpu_check	# family < 6, skip
+
+	andl	$0x000f00f0, %ecx	# mask model and extended model
+	shrl	$4, %ecx
+	cmpl	$0xd, %ecx
+	jb	verify_cpu_check	# family == 6, model < 0xd, skip
+
+verify_cpu_clear_xd:
+	movl	$MSR_IA32_MISC_ENABLE, %ecx
+	rdmsr
+	btrl	$2, %edx		# clear MSR_IA32_MISC_ENABLE_XD_DISABLE
+	jnc	verify_cpu_check	# only write MSR if bit was changed
+	wrmsr
+
+verify_cpu_check:
 	movl    $0x1,%eax		# Does the cpu have what it takes
 	cpuid
 	andl	$REQUIRED_MASK0,%edx
Signature · 1 line
1.7.2.3

A message that is not in the archive (sent off-list or not yet received): [email protected]

author
#4 lore raw

Re: [tip:x86/cpu] x86, cpu: Clear XD_DISABLED flag on Intel to regain NX

Quoted 8 lines · tip-bot for Kees Cook
Commit-ID:  ae84739c27b6b3725993202fe02ff35ab86468e1
Gitweb:     http://git.kernel.org/tip/ae84739c27b6b3725993202fe02ff35ab86468e1
Author:     Kees Cook <[email protected]>
AuthorDate: Wed, 10 Nov 2010 10:35:52 -0800
Committer:  H. Peter Anvin <[email protected]>
CommitDate: Wed, 10 Nov 2010 15:42:54 -0800

x86, cpu: Clear XD_DISABLED flag on Intel to regain NX

Just some quick positive feedback on this patch series: I've had at least one confirmed report of a user with an Asus system that had NX disabled (in the BIOS which offered no toggle to switch it on) is able to use the NX bit on their CPU finally. Additionally, no regressions have been identified either, and there has been a fair number of people running with the patch for a few weeks now in Ubuntu's development kernel.

-Kees

Signature · 2 lines
Kees Cook
Ubuntu Security Team
#5 lore raw

Re: [tip:x86/cpu] x86, cpu: Clear XD_DISABLED flag on Intel to regain NX

Quoted 8 lines · Kees Cook
Quoted 7 lines · tip-bot for Kees Cook
Commit-ID:  ae84739c27b6b3725993202fe02ff35ab86468e1
Gitweb:     http://git.kernel.org/tip/ae84739c27b6b3725993202fe02ff35ab86468e1
Author:     Kees Cook <[email protected]>
AuthorDate: Wed, 10 Nov 2010 10:35:52 -0800
Committer:  H. Peter Anvin <[email protected]>
CommitDate: Wed, 10 Nov 2010 15:42:54 -0800
Quoted 8 lines
Quoted 1 line

x86, cpu: Clear XD_DISABLED flag on Intel to regain NX

Just some quick positive feedback on this patch series: I've had at least one confirmed report of a user with an Asus system that had NX disabled (in the BIOS which offered no toggle to switch it on) is able to use the NX bit on their CPU finally. Additionally, no regressions have been identified either, and there has been a fair number of people running with the patch for a few weeks now in Ubuntu's development kernel.

Nice!

Ingo

author
#6 lore raw

PATCH v5 3/4 x86: call verify_cpu during 32bit CPU startup

In reply to Kees Cook

The XD_DISABLE-clearing side-effect needs to happen for both 32bit and 64bit, but the 32bit init routines were not calling verify_cpu() yet. This adds that call to gain the side-effect.

The longmode/SSE tests being performed in verify_cpu() need to happen very early for 64bit but not for 32bit. Instead of including it in two places for 32bit, we can just include it once in arch/x86/kernel/head_32.S.

---
 arch/x86/kernel/head_32.S    |    6 ++++++
 arch/x86/kernel/verify_cpu.S |    1 +
 2 files changed, 7 insertions(+), 0 deletions(-)
arch/x86/kernel/head_32.S +6 -0
index bcece91..fdaea52 100644
@@ -314,6 +314,10 @@ ENTRY(startup_32_smp)
 	subl $0x80000001, %eax
 	cmpl $(0x8000ffff-0x80000001), %eax
 	ja 6f
+
+	/* Clear bogus XD_DISABLE bits */
+	call verify_cpu
+
 	mov $0x80000001, %eax
 	cpuid
 	/* Execute Disable bit supported? */
@@ -609,6 +613,8 @@ ignore_int:
 #endif
 	iret
 
+#include "verify_cpu.S"
+
 	__REFDATA
 .align 4
 ENTRY(initial_code)
arch/x86/kernel/verify_cpu.S +1 -0
index ccb4136..5644b4b 100644
@@ -19,6 +19,7 @@
  *
  *	arch/x86_64/boot/compressed/head_64.S: Boot cpu verification
  *	arch/x86_64/kernel/trampoline_64.S: secondary processor verfication
+ *	arch/x86_64/kernel/head_32.S: processor startup
  *
  *	verify_cpu, returns the status of longmode and SSE in register %eax.
  *		0: Success    1: Failure
Signature · 1 line
1.7.2.3
#7 lore raw

Re: [PATCH 3/4] x86: call verify_cpu during 32bit CPU startup

Quoted 45 lines · Kees Cook

The XD_DISABLE-clearing side-effect needs to happen for both 32bit and 64bit, but the 32bit init routines were not calling verify_cpu() yet. This adds that call to gain the side-effect.

The longmode/SSE tests being performed in verify_cpu() need to happen very early for 64bit but not for 32bit. Instead of including it in two places for 32bit, we can just include it once in arch/x86/kernel/head_32.S.

---
 arch/x86/kernel/head_32.S    |    6 ++++++
 arch/x86/kernel/verify_cpu.S |    1 +

 2 files changed, 7 insertions(+), 0 deletions(-)
arch/x86/kernel/head_32.S +0 -0
index bcece91..fdaea52 100644
@@ -314,6 +314,10 @@ ENTRY(startup_32_smp)

       subl $0x80000001, %eax
       cmpl $(0x8000ffff-0x80000001), %eax
       ja 6f

arch/x86/kernel/head_32.S +4 -0
+
+       /* Clear bogus XD_DISABLE bits */
+       call verify_cpu
+

       mov $0x80000001, %eax
       cpuid
       /* Execute Disable bit supported? */

arch/x86/kernel/head_32.S +0 -0
@@ -609,6 +613,8 @@ ignore_int:
  #endif

       iret

arch/x86/kernel/head_32.S +2 -0
+#include "verify_cpu.S"
+

       __REFDATA
 .align 4
 ENTRY(initial_code)

arch/x86/kernel/verify_cpu.S +0 -0
index ccb4136..5644b4b 100644

arch/x86/kernel/verify_cpu.S line 22

Quoted 4 lines
arch/x86/kernel/verify_cpu.S +1 -0
@@ -19,6 +19,7 @@
  *
  *     arch/x86_64/boot/compressed/head_64.S: Boot cpu verification
  *     arch/x86_64/kernel/trampoline_64.S: secondary processor verfication
+ *     arch/x86_64/kernel/head_32.S: processor startup

should change those x86_64 to x86.

Yinghai

author
#8 lore raw

PATCH v5 4/4 x86: only CPU features determine NX capabilities

In reply to Kees Cook

Fix the NX feature boot warning when NX is missing to correctly reflect that BIOSes cannot disable NX now.

---
 arch/x86/mm/setup_nx.c |    2 +-
 1 files changed, 1 insertions(+), 1 deletions(-)
arch/x86/mm/setup_nx.c +1 -1
index a3250aa..410531d 100644
@@ -41,7 +41,7 @@ void __init x86_report_nx(void)
 {
 	if (!cpu_has_nx) {
 		printk(KERN_NOTICE "Notice: NX (Execute Disable) protection "
-		       "missing in CPU or disabled in BIOS!\n");
+		       "missing in CPU!\n");
 	} else {
 #if defined(CONFIG_X86_64) || defined(CONFIG_X86_PAE)
 		if (disable_nx) {
Signature · 1 line
1.7.2.3