Merge branch 'x86-mpx-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip

Pull x86 MPX fixes from Thomas Gleixner: "Three updates for the new MPX infrastructure: - Use the proper error check in the trap handler - Add a proper config option for it - Bring documentation up to date" * 'x86-mpx-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip: x86, mpx: Give MPX a real config option prompt x86, mpx: Update documentation x86_64/traps: Fix always true condition

Merge branch 'x86-mpx-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip
Pull x86 MPX fixes from Thomas Gleixner: "Three updates for the new MPX infrastructure: - Use the proper error check in the trap handler - Add a proper config option for it - Bring documentation up to date" * 'x86-mpx-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip: x86, mpx: Give MPX a real config option prompt x86, mpx: Update documentation x86_64/traps: Fix always true condition
Linus Torvalds
2 parents 1092b596a5 72e9b5fe9b
Showing 3 changed files Side-by-side Diff
Documentation/x86/intel_mpx.txt
arch/x86/Kconfig
arch/x86/kernel/traps.c
@@ -7,11 +7,15 @@
 references, for those references whose compile-time normal intentions are
 usurped at runtime due to buffer overflow or underflow.
  
+You can tell if your CPU supports MPX by looking in /proc/cpuinfo:
+
+	cat /proc/cpuinfo  | grep ' mpx '
+
 For more information, please refer to Intel(R) Architecture Instruction
 Set Extensions Programming Reference, Chapter 9: Intel(R) Memory Protection
 Extensions.
  
-Note: Currently no hardware with MPX ISA is available but it is always
+Note: As of December 2014, no hardware with MPX is available but it is
 possible to use SDE (Intel(R) Software Development Emulator) instead, which
 can be downloaded from
 http://software.intel.com/en-us/articles/intel-software-development-emulator
@@ -30,9 +34,15 @@
    instrumentation as well as some setup code called early after the app
    starts. New instruction prefixes are noops for old CPUs.
 2) That setup code allocates (virtual) space for the "bounds directory",
-   points the "bndcfgu" register to the directory and notifies the kernel
-   (via the new prctl(PR_MPX_ENABLE_MANAGEMENT)) that the app will be using
-   MPX.
+   points the "bndcfgu" register to the directory (must also set the valid
+   bit) and notifies the kernel (via the new prctl(PR_MPX_ENABLE_MANAGEMENT))
+   that the app will be using MPX.  The app must be careful not to access
+   the bounds tables between the time when it populates "bndcfgu" and
+   when it calls the prctl().  This might be hard to guarantee if the app
+   is compiled with MPX.  You can add "__attribute__((bnd_legacy))" to
+   the function to disable MPX instrumentation to help guarantee this.
+   Also be careful not to call out to any other code which might be
+   MPX-instrumented.
 3) The kernel detects that the CPU has MPX, allows the new prctl() to
    succeed, and notes the location of the bounds directory. Userspace is
    expected to keep the bounds directory at that locationWe note it
@@ -249,10 +249,6 @@
 	def_bool y
 	depends on INTEL_IOMMU && ACPI
  
-config X86_INTEL_MPX
-	def_bool y
-	depends on CPU_SUP_INTEL
-
 config X86_32_SMP
 	def_bool y
 	depends on X86_32 && SMP
@@ -1593,6 +1589,32 @@
 	  also a small increase in the kernel size if this is enabled.
  
 	  If unsure, say Y.
+
+config X86_INTEL_MPX
+	prompt "Intel MPX (Memory Protection Extensions)"
+	def_bool n
+	depends on CPU_SUP_INTEL
+	---help---
+	  MPX provides hardware features that can be used in
+	  conjunction with compiler-instrumented code to check
+	  memory references.  It is designed to detect buffer
+	  overflow or underflow bugs.
+
+	  This option enables running applications which are
+	  instrumented or otherwise use MPX.  It does not use MPX
+	  itself inside the kernel or to protect the kernel
+	  against bad memory references.
+
+	  Enabling this option will make the kernel larger:
+	  ~8k of kernel text and 36 bytes of data on a 64-bit
+	  defconfig.  It adds a long to the 'mm_struct' which
+	  will increase the kernel memory overhead of each
+	  process and adds some branches to paths used during
+	  exec() and munmap().
+
+	  For details, see Documentation/x86/intel_mpx.txt
+
+	  If unsure, say N.
  
 config EFI
 	bool "EFI runtime service support"
@@ -331,7 +331,7 @@
 		break; /* Success, it was handled */
 	case 1: /* Bound violation. */
 		info = mpx_generate_siginfo(regs, xsave_buf);
-		if (PTR_ERR(info)) {
+		if (IS_ERR(info)) {
 			/*
 			 * We failed to decode the MPX instruction.  Act as if
 			 * the exception was not caused by MPX.
...	...	@@ -7,11 +7,15 @@
7	7	references, for those references whose compile-time normal intentions are
8	8	usurped at runtime due to buffer overflow or underflow.
9	9
	10	+You can tell if your CPU supports MPX by looking in /proc/cpuinfo:
	11	+
	12	+ cat /proc/cpuinfo \| grep ' mpx '
	13	+
10	14	For more information, please refer to Intel(R) Architecture Instruction
11	15	Set Extensions Programming Reference, Chapter 9: Intel(R) Memory Protection
12	16	Extensions.
13	17
14		-Note: Currently no hardware with MPX ISA is available but it is always
	18	+Note: As of December 2014, no hardware with MPX is available but it is
15	19	possible to use SDE (Intel(R) Software Development Emulator) instead, which
16	20	can be downloaded from
17	21	http://software.intel.com/en-us/articles/intel-software-development-emulator
...	...	@@ -30,9 +34,15 @@
30	34	instrumentation as well as some setup code called early after the app
31	35	starts. New instruction prefixes are noops for old CPUs.
32	36	2) That setup code allocates (virtual) space for the "bounds directory",
33		- points the "bndcfgu" register to the directory and notifies the kernel
34		- (via the new prctl(PR_MPX_ENABLE_MANAGEMENT)) that the app will be using
35		- MPX.
	37	+ points the "bndcfgu" register to the directory (must also set the valid
	38	+ bit) and notifies the kernel (via the new prctl(PR_MPX_ENABLE_MANAGEMENT))
	39	+ that the app will be using MPX. The app must be careful not to access
	40	+ the bounds tables between the time when it populates "bndcfgu" and
	41	+ when it calls the prctl(). This might be hard to guarantee if the app
	42	+ is compiled with MPX. You can add "__attribute__((bnd_legacy))" to
	43	+ the function to disable MPX instrumentation to help guarantee this.
	44	+ Also be careful not to call out to any other code which might be
	45	+ MPX-instrumented.
36	46	3) The kernel detects that the CPU has MPX, allows the new prctl() to
37	47	succeed, and notes the location of the bounds directory. Userspace is
38	48	expected to keep the bounds directory at that locationWe note it
...	...	@@ -249,10 +249,6 @@
249	249	def_bool y
250	250	depends on INTEL_IOMMU && ACPI
251	251
252		-config X86_INTEL_MPX
253		- def_bool y
254		- depends on CPU_SUP_INTEL
255		-
256	252	config X86_32_SMP
257	253	def_bool y
258	254	depends on X86_32 && SMP
...	...	@@ -1593,6 +1589,32 @@
1593	1589	also a small increase in the kernel size if this is enabled.
1594	1590
1595	1591	If unsure, say Y.
	1592	+
	1593	+config X86_INTEL_MPX
	1594	+ prompt "Intel MPX (Memory Protection Extensions)"
	1595	+ def_bool n
	1596	+ depends on CPU_SUP_INTEL
	1597	+ ---help---
	1598	+ MPX provides hardware features that can be used in
	1599	+ conjunction with compiler-instrumented code to check
	1600	+ memory references. It is designed to detect buffer
	1601	+ overflow or underflow bugs.
	1602	+
	1603	+ This option enables running applications which are
	1604	+ instrumented or otherwise use MPX. It does not use MPX
	1605	+ itself inside the kernel or to protect the kernel
	1606	+ against bad memory references.
	1607	+
	1608	+ Enabling this option will make the kernel larger:
	1609	+ ~8k of kernel text and 36 bytes of data on a 64-bit
	1610	+ defconfig. It adds a long to the 'mm_struct' which
	1611	+ will increase the kernel memory overhead of each
	1612	+ process and adds some branches to paths used during
	1613	+ exec() and munmap().
	1614	+
	1615	+ For details, see Documentation/x86/intel_mpx.txt
	1616	+
	1617	+ If unsure, say N.
1596	1618
1597	1619	config EFI
1598	1620	bool "EFI runtime service support"
...	...	@@ -331,7 +331,7 @@
331	331	break; /* Success, it was handled */
332	332	case 1: /* Bound violation. */
333	333	info = mpx_generate_siginfo(regs, xsave_buf);
334		- if (PTR_ERR(info)) {
	334	+ if (IS_ERR(info)) {
335	335	/*
336	336	* We failed to decode the MPX instruction. Act as if
337	337	* the exception was not caused by MPX.