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

Pull x86 mm updates from Ingo Molnar:
 "A laundry list of changes: KASAN improvements/fixes for ptdump, a
  self-test fix, PAT cleanup and wbinvd() avoidance, removal of stale
  code and documentation updates"

* 'x86-mm-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
  x86/mm/ptdump: Add address marker for KASAN shadow region
  x86/mm/ptdump: Optimize check for W+X mappings for CONFIG_KASAN=y
  x86/mm/pat: Use rb_entry()
  x86/mpx: Re-add MPX to selftests Makefile
  x86/mm: Remove CONFIG_DEBUG_NX_TEST
  x86/mm/cpa: Avoid wbinvd() for PREEMPT
  x86/mm: Improve documentation for low-level device I/O functions

arch/x86/Kconfig.debug

@@ -120,14 +120,6 @@ config DEBUG_SET_MODULE_RONX
 	  against certain classes of kernel exploits.
 	  If in doubt, say "N".
 
-config DEBUG_NX_TEST
-	tristate "Testcase for the NX non-executable stack feature"
-	depends on DEBUG_KERNEL && m
-	---help---
-	  This option enables a testcase for the CPU NX capability
-	  and the software setup of this feature.
-	  If in doubt, say "N"
-
 config DOUBLEFAULT
 	default y
 	bool "Enable doublefault exception handler" if EXPERT

arch/x86/include/asm/io.h

@@ -164,6 +164,17 @@ static inline unsigned int isa_virt_to_bus(volatile void *address)
 #define virt_to_bus virt_to_phys
 #define bus_to_virt phys_to_virt
 
+/*
+ * The default ioremap() behavior is non-cached; if you need something
+ * else, you probably want one of the following.
+ */
+extern void __iomem *ioremap_nocache(resource_size_t offset, unsigned long size);
+extern void __iomem *ioremap_uc(resource_size_t offset, unsigned long size);
+#define ioremap_uc ioremap_uc
+
+extern void __iomem *ioremap_cache(resource_size_t offset, unsigned long size);
+extern void __iomem *ioremap_prot(resource_size_t offset, unsigned long size, unsigned long prot_val);
+
 /**
  * ioremap     -   map bus memory into CPU space
  * @offset:    bus address of the memory
@@ -178,17 +189,6 @@ static inline unsigned int isa_virt_to_bus(volatile void *address)
  * If the area you are trying to map is a PCI BAR you should have a
  * look at pci_iomap().
  */
-extern void __iomem *ioremap_nocache(resource_size_t offset, unsigned long size);
-extern void __iomem *ioremap_uc(resource_size_t offset, unsigned long size);
-#define ioremap_uc ioremap_uc
-
-extern void __iomem *ioremap_cache(resource_size_t offset, unsigned long size);
-extern void __iomem *ioremap_prot(resource_size_t offset, unsigned long size,
-				unsigned long prot_val);
-
-/*
- * The default ioremap() behavior is non-cached:
- */
 static inline void __iomem *ioremap(resource_size_t offset, unsigned long size)
 {
 	return ioremap_nocache(offset, size);
@@ -207,18 +207,42 @@ extern void set_iounmap_nonlazy(void);
  */
 #define xlate_dev_kmem_ptr(p)	p
 
+/**
+ * memset_io	Set a range of I/O memory to a constant value
+ * @addr:	The beginning of the I/O-memory range to set
+ * @val:	The value to set the memory to
+ * @count:	The number of bytes to set
+ *
+ * Set a range of I/O memory to a given value.
+ */
 static inline void
 memset_io(volatile void __iomem *addr, unsigned char val, size_t count)
 {
 	memset((void __force *)addr, val, count);
 }
 
+/**
+ * memcpy_fromio	Copy a block of data from I/O memory
+ * @dst:		The (RAM) destination for the copy
+ * @src:		The (I/O memory) source for the data
+ * @count:		The number of bytes to copy
+ *
+ * Copy a block of data from I/O memory.
+ */
 static inline void
 memcpy_fromio(void *dst, const volatile void __iomem *src, size_t count)
 {
 	memcpy(dst, (const void __force *)src, count);
 }
 
+/**
+ * memcpy_toio		Copy a block of data into I/O memory
+ * @dst:		The (I/O memory) destination for the copy
+ * @src:		The (RAM) source for the data
+ * @count:		The number of bytes to copy
+ *
+ * Copy a block of data to I/O memory.
+ */
 static inline void
 memcpy_toio(volatile void __iomem *dst, const void *src, size_t count)
 {

arch/x86/kernel/Makefile

@@ -101,7 +101,6 @@ obj-$(CONFIG_APB_TIMER)		+= apb_timer.o
 
 obj-$(CONFIG_AMD_NB)		+= amd_nb.o
 obj-$(CONFIG_DEBUG_RODATA_TEST)	+= test_rodata.o
-obj-$(CONFIG_DEBUG_NX_TEST)	+= test_nx.o
 obj-$(CONFIG_DEBUG_NMI_SELFTEST) += nmi_selftest.o
 
 obj-$(CONFIG_KVM_GUEST)		+= kvm.o kvmclock.o

arch/x86/kernel/test_nx.c

-/*
- * test_nx.c: functional test for NX functionality
- *
- * (C) Copyright 2008 Intel Corporation
- * Author: Arjan van de Ven <arjan@linux.intel.com>
- *
- * This program is free software; you can redistribute it and/or
- * modify it under the terms of the GNU General Public License
- * as published by the Free Software Foundation; version 2
- * of the License.
- */
-#include <linux/module.h>
-#include <linux/sort.h>
-#include <linux/slab.h>
-
-#include <linux/uaccess.h>
-#include <asm/asm.h>
-
-extern int rodata_test_data;
-
-/*
- * This file checks 4 things:
- * 1) Check if the stack is not executable
- * 2) Check if kmalloc memory is not executable
- * 3) Check if the .rodata section is not executable
- * 4) Check if the .data section of a module is not executable
- *
- * To do this, the test code tries to execute memory in stack/kmalloc/etc,
- * and then checks if the expected trap happens.
- *
- * Sadly, this implies having a dynamic exception handling table entry.
- * ... which can be done (and will make Rusty cry)... but it can only
- * be done in a stand-alone module with only 1 entry total.
- * (otherwise we'd have to sort and that's just too messy)
- */
-
-
-
-/*
- * We want to set up an exception handling point on our stack,
- * which means a variable value. This function is rather dirty
- * and walks the exception table of the module, looking for a magic
- * marker and replaces it with a specific function.
- */
-static void fudze_exception_table(void *marker, void *new)
-{
-	struct module *mod = THIS_MODULE;
-	struct exception_table_entry *extable;
-
-	/*
-	 * Note: This module has only 1 exception table entry,
-	 * so searching and sorting is not needed. If that changes,
-	 * this would be the place to search and re-sort the exception
-	 * table.
-	 */
-	if (mod->num_exentries > 1) {
-		printk(KERN_ERR "test_nx: too many exception table entries!\n");
-		printk(KERN_ERR "test_nx: test results are not reliable.\n");
-		return;
-	}
-	extable = (struct exception_table_entry *)mod->extable;
-	extable[0].insn = (unsigned long)new;
-}
-
-
-/*
- * exception tables get their symbols translated so we need
- * to use a fake function to put in there, which we can then
- * replace at runtime.
- */
-void foo_label(void);
-
-/*
- * returns 0 for not-executable, negative for executable
- *
- * Note: we cannot allow this function to be inlined, because
- * that would give us more than 1 exception table entry.
- * This in turn would break the assumptions above.
- */
-static noinline int test_address(void *address)
-{
-	unsigned long result;
-
-	/* Set up an exception table entry for our address */
-	fudze_exception_table(&foo_label, address);
-	result = 1;
-	asm volatile(
-		"foo_label:\n"
-		"0:	call *%[fake_code]\n"
-		"1:\n"
-		".section .fixup,\"ax\"\n"
-		"2:	mov %[zero], %[rslt]\n"
-		"	ret\n"
-		".previous\n"
-		_ASM_EXTABLE(0b,2b)
-		: [rslt] "=r" (result)
-		: [fake_code] "r" (address), [zero] "r" (0UL), "0" (result)
-	);
-	/* change the exception table back for the next round */
-	fudze_exception_table(address, &foo_label);
-
-	if (result)
-		return -ENODEV;
-	return 0;
-}
-
-static unsigned char test_data = 0xC3; /* 0xC3 is the opcode for "ret" */
-
-static int test_NX(void)
-{
-	int ret = 0;
-	/* 0xC3 is the opcode for "ret" */
-	char stackcode[] = {0xC3, 0x90, 0 };
-	char *heap;
-
-	test_data = 0xC3;
-
-	printk(KERN_INFO "Testing NX protection\n");
-
-	/* Test 1: check if the stack is not executable */
-	if (test_address(&stackcode)) {
-		printk(KERN_ERR "test_nx: stack was executable\n");
-		ret = -ENODEV;
-	}
-
-
-	/* Test 2: Check if the heap is executable */
-	heap = kmalloc(64, GFP_KERNEL);
-	if (!heap)
-		return -ENOMEM;
-	heap[0] = 0xC3; /* opcode for "ret" */
-
-	if (test_address(heap)) {
-		printk(KERN_ERR "test_nx: heap was executable\n");
-		ret = -ENODEV;
-	}
-	kfree(heap);
-
-	/*
-	 * The following 2 tests currently fail, this needs to get fixed
-	 * Until then, don't run them to avoid too many people getting scared
-	 * by the error message
-	 */
-
-	/* Test 3: Check if the .rodata section is executable */
-	if (rodata_test_data != 0xC3) {
-		printk(KERN_ERR "test_nx: .rodata marker has invalid value\n");
-		ret = -ENODEV;
-	} else if (test_address(&rodata_test_data)) {
-		printk(KERN_ERR "test_nx: .rodata section is executable\n");
-		ret = -ENODEV;
-	}
-
-#if 0
-	/* Test 4: Check if the .data section of a module is executable */
-	if (test_address(&test_data)) {
-		printk(KERN_ERR "test_nx: .data section is executable\n");
-		ret = -ENODEV;
-	}
-
-#endif
-	return ret;
-}
-
-static void test_exit(void)
-{
-}
-
-module_init(test_NX);
-module_exit(test_exit);
-MODULE_LICENSE("GPL");
-MODULE_DESCRIPTION("Testcase for the NX infrastructure");
-MODULE_AUTHOR("Arjan van de Ven <arjan@linux.intel.com>");

arch/x86/mm/dump_pagetables.c

@@ -18,6 +18,7 @@
 #include <linux/sched.h>
 #include <linux/seq_file.h>
 
+#include <asm/kasan.h>
 #include <asm/pgtable.h>
 
 /*
@@ -51,6 +52,10 @@ enum address_markers_idx {
 	LOW_KERNEL_NR,
 	VMALLOC_START_NR,
 	VMEMMAP_START_NR,
+#ifdef CONFIG_KASAN
+	KASAN_SHADOW_START_NR,
+	KASAN_SHADOW_END_NR,
+#endif
 # ifdef CONFIG_X86_ESPFIX64
 	ESPFIX_START_NR,
 # endif
@@ -76,6 +81,10 @@ static struct addr_marker address_markers[] = {
 	{ 0/* PAGE_OFFSET */,   "Low Kernel Mapping" },
 	{ 0/* VMALLOC_START */, "vmalloc() Area" },
 	{ 0/* VMEMMAP_START */, "Vmemmap" },
+#ifdef CONFIG_KASAN
+	{ KASAN_SHADOW_START,	"KASAN shadow" },
+	{ KASAN_SHADOW_END,	"KASAN shadow end" },
+#endif
 # ifdef CONFIG_X86_ESPFIX64
 	{ ESPFIX_BASE_ADDR,	"ESPfix Area", 16 },
 # endif
@@ -327,18 +336,31 @@ static void walk_pmd_level(struct seq_file *m, struct pg_state *st, pud_t addr,
 
 #if PTRS_PER_PUD > 1
 
+/*
+ * This is an optimization for CONFIG_DEBUG_WX=y + CONFIG_KASAN=y
+ * KASAN fills page tables with the same values. Since there is no
+ * point in checking page table more than once we just skip repeated
+ * entries. This saves us dozens of seconds during boot.
+ */
+static bool pud_already_checked(pud_t *prev_pud, pud_t *pud, bool checkwx)
+{
+	return checkwx && prev_pud && (pud_val(*prev_pud) == pud_val(*pud));
+}
+
 static void walk_pud_level(struct seq_file *m, struct pg_state *st, pgd_t addr,
 							unsigned long P)
 {
 	int i;
 	pud_t *start;
 	pgprotval_t prot;
+	pud_t *prev_pud = NULL;
 
 	start = (pud_t *) pgd_page_vaddr(addr);
 
 	for (i = 0; i < PTRS_PER_PUD; i++) {
 		st->current_address = normalize_addr(P + i * PUD_LEVEL_MULT);
-		if (!pud_none(*start)) {
+		if (!pud_none(*start) &&
+		    !pud_already_checked(prev_pud, start, st->check_wx)) {
 			if (pud_large(*start) || !pud_present(*start)) {
 				prot = pud_flags(*start);
 				note_page(m, st, __pgprot(prot), 2);
@@ -349,6 +371,7 @@ static void walk_pud_level(struct seq_file *m, struct pg_state *st, pgd_t addr,
 		} else
 			note_page(m, st, __pgprot(0), 2);
 
+		prev_pud = start;
 		start++;
 	}
 }

arch/x86/mm/pageattr.c

@@ -214,7 +214,20 @@ static void cpa_flush_array(unsigned long *start, int numpages, int cache,
 			    int in_flags, struct page **pages)
 {
 	unsigned int i, level;
+#ifdef CONFIG_PREEMPT
+	/*
+	 * Avoid wbinvd() because it causes latencies on all CPUs,
+	 * regardless of any CPU isolation that may be in effect.
+	 *
+	 * This should be extended for CAT enabled systems independent of
+	 * PREEMPT because wbinvd() does not respect the CAT partitions and
+	 * this is exposed to unpriviledged users through the graphics
+	 * subsystem.
+	 */
+	unsigned long do_wbinvd = 0;
+#else
 	unsigned long do_wbinvd = cache && numpages >= 1024; /* 4M threshold */
+#endif
 
 	BUG_ON(irqs_disabled());
 

arch/x86/mm/pat_rbtree.c

@@ -47,7 +47,7 @@ static u64 get_subtree_max_end(struct rb_node *node)
 {
 	u64 ret = 0;
 	if (node) {
-		struct memtype *data = container_of(node, struct memtype, rb);
+		struct memtype *data = rb_entry(node, struct memtype, rb);
 		ret = data->subtree_max_end;
 	}
 	return ret;
@@ -79,7 +79,7 @@ static struct memtype *memtype_rb_lowest_match(struct rb_root *root,
 	struct memtype *last_lower = NULL;
 
 	while (node) {
-		struct memtype *data = container_of(node, struct memtype, rb);
+		struct memtype *data = rb_entry(node, struct memtype, rb);
 
 		if (get_subtree_max_end(node->rb_left) > start) {
 			/* Lowest overlap if any must be on left side */
@@ -121,7 +121,7 @@ static struct memtype *memtype_rb_match(struct rb_root *root,
 
 		node = rb_next(&match->rb);
 		if (node)
-			match = container_of(node, struct memtype, rb);
+			match = rb_entry(node, struct memtype, rb);
 		else
 			match = NULL;
 	}
@@ -150,7 +150,7 @@ static int memtype_rb_check_conflict(struct rb_root *root,
 
 	node = rb_next(&match->rb);
 	while (node) {
-		match = container_of(node, struct memtype, rb);
+		match = rb_entry(node, struct memtype, rb);
 
 		if (match->start >= end) /* Checked all possible matches */
 			goto success;
@@ -181,7 +181,7 @@ static void memtype_rb_insert(struct rb_root *root, struct memtype *newdata)
 	struct rb_node *parent = NULL;
 
 	while (*node) {
-		struct memtype *data = container_of(*node, struct memtype, rb);
+		struct memtype *data = rb_entry(*node, struct memtype, rb);
 
 		parent = *node;
 		if (data->subtree_max_end < newdata->end)
@@ -270,7 +270,7 @@ int rbt_memtype_copy_nth_element(struct memtype *out, loff_t pos)
 	}
 
 	if (node) { /* pos == i */
-		struct memtype *this = container_of(node, struct memtype, rb);
+		struct memtype *this = rb_entry(node, struct memtype, rb);
 		*out = *this;
 		return 0;
 	} else {

tools/testing/selftests/x86/Makefile

@@ -5,7 +5,7 @@ include ../lib.mk
 .PHONY: all all_32 all_64 warn_32bit_failure clean
 
 TARGETS_C_BOTHBITS := single_step_syscall sysret_ss_attrs syscall_nt ptrace_syscall test_mremap_vdso \
-			check_initial_reg_state sigreturn ldt_gdt iopl \
+			check_initial_reg_state sigreturn ldt_gdt iopl mpx-mini-test \
 			protection_keys test_vdso
 TARGETS_C_32BIT_ONLY := entry_from_vm86 syscall_arg_fault test_syscall_vdso unwind_vdso \
 			test_FCMOV test_FCOMI test_FISTTP \