Merge commit 'perf/core' into perf/hw-breakpoint

Conflicts:
	kernel/Makefile
	kernel/trace/Makefile
	kernel/trace/trace.h
	samples/Makefile

Merge reason: We need to be uptodate with the perf events development
branch because we plan to rewrite the breakpoints API on top of
perf events.

15 files changed, 437 insertions, 249 deletions

diff --git a/arch/x86/mm/Makefile b/arch/x86/mm/Makefile
index 9b5a9f5..06630d2 100644
--- a/arch/x86/mm/Makefile
+++ b/arch/x86/mm/Makefile
@@ -1,9 +1,10 @@
 obj-y	:=  init.o init_$(BITS).o fault.o ioremap.o extable.o pageattr.o mmap.o \
-	    pat.o pgtable.o physaddr.o gup.o
+	    pat.o pgtable.o physaddr.o gup.o setup_nx.o
 
 # Make sure __phys_addr has no stackprotector
 nostackp := $(call cc-option, -fno-stack-protector)
 CFLAGS_physaddr.o		:= $(nostackp)
+CFLAGS_setup_nx.o		:= $(nostackp)
 
 obj-$(CONFIG_SMP)		+= tlb.o
 
diff --git a/arch/x86/mm/fault.c b/arch/x86/mm/fault.c
index 775a020..f4cee90 100644
--- a/arch/x86/mm/fault.c
+++ b/arch/x86/mm/fault.c
@@ -10,7 +10,7 @@
 #include <linux/bootmem.h>		/* max_low_pfn			*/
 #include <linux/kprobes.h>		/* __kprobes, ...		*/
 #include <linux/mmiotrace.h>		/* kmmio_handler, ...		*/
-#include <linux/perf_counter.h>		/* perf_swcounter_event		*/
+#include <linux/perf_event.h>		/* perf_sw_event		*/
 
 #include <asm/traps.h>			/* dotraplinkage, ...		*/
 #include <asm/pgalloc.h>		/* pgd_*(), ...			*/
@@ -167,6 +167,7 @@ force_sig_info_fault(int si_signo, int si_code, unsigned long address,
 	info.si_errno	= 0;
 	info.si_code	= si_code;
 	info.si_addr	= (void __user *)address;
+	info.si_addr_lsb = si_code == BUS_MCEERR_AR ? PAGE_SHIFT : 0;
 
 	force_sig_info(si_signo, &info, tsk);
 }
@@ -790,10 +791,12 @@ out_of_memory(struct pt_regs *regs, unsigned long error_code,
 }
 
 static void
-do_sigbus(struct pt_regs *regs, unsigned long error_code, unsigned long address)
+do_sigbus(struct pt_regs *regs, unsigned long error_code, unsigned long address,
+	  unsigned int fault)
 {
 	struct task_struct *tsk = current;
 	struct mm_struct *mm = tsk->mm;
+	int code = BUS_ADRERR;
 
 	up_read(&mm->mmap_sem);
 
@@ -809,7 +812,15 @@ do_sigbus(struct pt_regs *regs, unsigned long error_code, unsigned long address)
 	tsk->thread.error_code	= error_code;
 	tsk->thread.trap_no	= 14;
 
-	force_sig_info_fault(SIGBUS, BUS_ADRERR, address, tsk);
+#ifdef CONFIG_MEMORY_FAILURE
+	if (fault & VM_FAULT_HWPOISON) {
+		printk(KERN_ERR
+	"MCE: Killing %s:%d due to hardware memory corruption fault at %lx\n",
+			tsk->comm, tsk->pid, address);
+		code = BUS_MCEERR_AR;
+	}
+#endif
+	force_sig_info_fault(SIGBUS, code, address, tsk);
 }
 
 static noinline void
@@ -819,8 +830,8 @@ mm_fault_error(struct pt_regs *regs, unsigned long error_code,
 	if (fault & VM_FAULT_OOM) {
 		out_of_memory(regs, error_code, address);
 	} else {
-		if (fault & VM_FAULT_SIGBUS)
-			do_sigbus(regs, error_code, address);
+		if (fault & (VM_FAULT_SIGBUS|VM_FAULT_HWPOISON))
+			do_sigbus(regs, error_code, address, fault);
 		else
 			BUG();
 	}
@@ -1017,7 +1028,7 @@ do_page_fault(struct pt_regs *regs, unsigned long error_code)
 	if (unlikely(error_code & PF_RSVD))
 		pgtable_bad(regs, error_code, address);
 
-	perf_swcounter_event(PERF_COUNT_SW_PAGE_FAULTS, 1, 0, regs, address);
+	perf_sw_event(PERF_COUNT_SW_PAGE_FAULTS, 1, 0, regs, address);
 
 	/*
 	 * If we're in an interrupt, have no user context or are running
@@ -1114,11 +1125,11 @@ good_area:
 
 	if (fault & VM_FAULT_MAJOR) {
 		tsk->maj_flt++;
-		perf_swcounter_event(PERF_COUNT_SW_PAGE_FAULTS_MAJ, 1, 0,
+		perf_sw_event(PERF_COUNT_SW_PAGE_FAULTS_MAJ, 1, 0,
 				     regs, address);
 	} else {
 		tsk->min_flt++;
-		perf_swcounter_event(PERF_COUNT_SW_PAGE_FAULTS_MIN, 1, 0,
+		perf_sw_event(PERF_COUNT_SW_PAGE_FAULTS_MIN, 1, 0,
 				     regs, address);
 	}
 
diff --git a/arch/x86/mm/init.c b/arch/x86/mm/init.c
index 0607119..73ffd55 100644
--- a/arch/x86/mm/init.c
+++ b/arch/x86/mm/init.c
@@ -28,69 +28,6 @@ int direct_gbpages
 #endif
 ;
 
-int nx_enabled;
-
-#if defined(CONFIG_X86_64) || defined(CONFIG_X86_PAE)
-static int disable_nx __cpuinitdata;
-
-/*
- * noexec = on|off
- *
- * Control non-executable mappings for processes.
- *
- * on      Enable
- * off     Disable
- */
-static int __init noexec_setup(char *str)
-{
-	if (!str)
-		return -EINVAL;
-	if (!strncmp(str, "on", 2)) {
-		__supported_pte_mask |= _PAGE_NX;
-		disable_nx = 0;
-	} else if (!strncmp(str, "off", 3)) {
-		disable_nx = 1;
-		__supported_pte_mask &= ~_PAGE_NX;
-	}
-	return 0;
-}
-early_param("noexec", noexec_setup);
-#endif
-
-#ifdef CONFIG_X86_PAE
-static void __init set_nx(void)
-{
-	unsigned int v[4], l, h;
-
-	if (cpu_has_pae && (cpuid_eax(0x80000000) > 0x80000001)) {
-		cpuid(0x80000001, &v[0], &v[1], &v[2], &v[3]);
-
-		if ((v[3] & (1 << 20)) && !disable_nx) {
-			rdmsr(MSR_EFER, l, h);
-			l |= EFER_NX;
-			wrmsr(MSR_EFER, l, h);
-			nx_enabled = 1;
-			__supported_pte_mask |= _PAGE_NX;
-		}
-	}
-}
-#else
-static inline void set_nx(void)
-{
-}
-#endif
-
-#ifdef CONFIG_X86_64
-void __cpuinit check_efer(void)
-{
-	unsigned long efer;
-
-	rdmsrl(MSR_EFER, efer);
-	if (!(efer & EFER_NX) || disable_nx)
-		__supported_pte_mask &= ~_PAGE_NX;
-}
-#endif
-
 static void __init find_early_table_space(unsigned long end, int use_pse,
 					  int use_gbpages)
 {
diff --git a/arch/x86/mm/init_32.c b/arch/x86/mm/init_32.c
index 3cd7711..30938c1 100644
--- a/arch/x86/mm/init_32.c
+++ b/arch/x86/mm/init_32.c
@@ -84,7 +84,7 @@ static pmd_t * __init one_md_table_init(pgd_t *pgd)
 #ifdef CONFIG_X86_PAE
 	if (!(pgd_val(*pgd) & _PAGE_PRESENT)) {
 		if (after_bootmem)
-			pmd_table = (pmd_t *)alloc_bootmem_low_pages(PAGE_SIZE);
+			pmd_table = (pmd_t *)alloc_bootmem_pages(PAGE_SIZE);
 		else
 			pmd_table = (pmd_t *)alloc_low_page();
 		paravirt_alloc_pmd(&init_mm, __pa(pmd_table) >> PAGE_SHIFT);
@@ -116,7 +116,7 @@ static pte_t * __init one_page_table_init(pmd_t *pmd)
 #endif
 			if (!page_table)
 				page_table =
-				(pte_t *)alloc_bootmem_low_pages(PAGE_SIZE);
+				(pte_t *)alloc_bootmem_pages(PAGE_SIZE);
 		} else
 			page_table = (pte_t *)alloc_low_page();
 
@@ -857,8 +857,6 @@ static void __init test_wp_bit(void)
 	}
 }
 
-static struct kcore_list kcore_mem, kcore_vmalloc;
-
 void __init mem_init(void)
 {
 	int codesize, reservedpages, datasize, initsize;
@@ -886,13 +884,9 @@ void __init mem_init(void)
 	datasize =  (unsigned long) &_edata - (unsigned long) &_etext;
 	initsize =  (unsigned long) &__init_end - (unsigned long) &__init_begin;
 
-	kclist_add(&kcore_mem, __va(0), max_low_pfn << PAGE_SHIFT);
-	kclist_add(&kcore_vmalloc, (void *)VMALLOC_START,
-		   VMALLOC_END-VMALLOC_START);
-
 	printk(KERN_INFO "Memory: %luk/%luk available (%dk kernel code, "
 			"%dk reserved, %dk data, %dk init, %ldk highmem)\n",
-		(unsigned long) nr_free_pages() << (PAGE_SHIFT-10),
+		nr_free_pages() << (PAGE_SHIFT-10),
 		num_physpages << (PAGE_SHIFT-10),
 		codesize >> 10,
 		reservedpages << (PAGE_SHIFT-10),
diff --git a/arch/x86/mm/init_64.c b/arch/x86/mm/init_64.c
index ea56b8c..5a4398a 100644
--- a/arch/x86/mm/init_64.c
+++ b/arch/x86/mm/init_64.c
@@ -647,8 +647,7 @@ EXPORT_SYMBOL_GPL(memory_add_physaddr_to_nid);
 
 #endif /* CONFIG_MEMORY_HOTPLUG */
 
-static struct kcore_list kcore_mem, kcore_vmalloc, kcore_kernel,
-			 kcore_modules, kcore_vsyscall;
+static struct kcore_list kcore_vsyscall;
 
 void __init mem_init(void)
 {
@@ -677,17 +676,12 @@ void __init mem_init(void)
 	initsize =  (unsigned long) &__init_end - (unsigned long) &__init_begin;
 
 	/* Register memory areas for /proc/kcore */
-	kclist_add(&kcore_mem, __va(0), max_low_pfn << PAGE_SHIFT);
-	kclist_add(&kcore_vmalloc, (void *)VMALLOC_START,
-		   VMALLOC_END-VMALLOC_START);
-	kclist_add(&kcore_kernel, &_stext, _end - _stext);
-	kclist_add(&kcore_modules, (void *)MODULES_VADDR, MODULES_LEN);
 	kclist_add(&kcore_vsyscall, (void *)VSYSCALL_START,
-				 VSYSCALL_END - VSYSCALL_START);
+			 VSYSCALL_END - VSYSCALL_START, KCORE_OTHER);
 
 	printk(KERN_INFO "Memory: %luk/%luk available (%ldk kernel code, "
 			 "%ldk absent, %ldk reserved, %ldk data, %ldk init)\n",
-		(unsigned long) nr_free_pages() << (PAGE_SHIFT-10),
+		nr_free_pages() << (PAGE_SHIFT-10),
 		max_pfn << (PAGE_SHIFT-10),
 		codesize >> 10,
 		absent_pages << (PAGE_SHIFT-10),
diff --git a/arch/x86/mm/iomap_32.c b/arch/x86/mm/iomap_32.c
index fe6f84c..84e236c 100644
--- a/arch/x86/mm/iomap_32.c
+++ b/arch/x86/mm/iomap_32.c
@@ -21,7 +21,7 @@
 #include <linux/module.h>
 #include <linux/highmem.h>
 
-int is_io_mapping_possible(resource_size_t base, unsigned long size)
+static int is_io_mapping_possible(resource_size_t base, unsigned long size)
 {
 #if !defined(CONFIG_X86_PAE) && defined(CONFIG_PHYS_ADDR_T_64BIT)
 	/* There is no way to map greater than 1 << 32 address without PAE */
@@ -30,7 +30,30 @@ int is_io_mapping_possible(resource_size_t base, unsigned long size)
 #endif
 	return 1;
 }
-EXPORT_SYMBOL_GPL(is_io_mapping_possible);
+
+int iomap_create_wc(resource_size_t base, unsigned long size, pgprot_t *prot)
+{
+	unsigned long flag = _PAGE_CACHE_WC;
+	int ret;
+
+	if (!is_io_mapping_possible(base, size))
+		return -EINVAL;
+
+	ret = io_reserve_memtype(base, base + size, &flag);
+	if (ret)
+		return ret;
+
+	*prot = __pgprot(__PAGE_KERNEL | flag);
+	return 0;
+}
+EXPORT_SYMBOL_GPL(iomap_create_wc);
+
+void
+iomap_free(resource_size_t base, unsigned long size)
+{
+	io_free_memtype(base, base + size);
+}
+EXPORT_SYMBOL_GPL(iomap_free);
 
 void *kmap_atomic_prot_pfn(unsigned long pfn, enum km_type type, pgprot_t prot)
 {
diff --git a/arch/x86/mm/ioremap.c b/arch/x86/mm/ioremap.c
index 04e1ad6..334e63c 100644
--- a/arch/x86/mm/ioremap.c
+++ b/arch/x86/mm/ioremap.c
@@ -158,24 +158,14 @@ static void __iomem *__ioremap_caller(resource_size_t phys_addr,
 	retval = reserve_memtype(phys_addr, (u64)phys_addr + size,
 						prot_val, &new_prot_val);
 	if (retval) {
-		pr_debug("Warning: reserve_memtype returned %d\n", retval);
+		printk(KERN_ERR "ioremap reserve_memtype failed %d\n", retval);
 		return NULL;
 	}
 
 	if (prot_val != new_prot_val) {
-		/*
-		 * Do not fallback to certain memory types with certain
-		 * requested type:
-		 * - request is uc-, return cannot be write-back
-		 * - request is uc-, return cannot be write-combine
-		 * - request is write-combine, return cannot be write-back
-		 */
-		if ((prot_val == _PAGE_CACHE_UC_MINUS &&
-		     (new_prot_val == _PAGE_CACHE_WB ||
-		      new_prot_val == _PAGE_CACHE_WC)) ||
-		    (prot_val == _PAGE_CACHE_WC &&
-		     new_prot_val == _PAGE_CACHE_WB)) {
-			pr_debug(
+		if (!is_new_memtype_allowed(phys_addr, size,
+					    prot_val, new_prot_val)) {
+			printk(KERN_ERR
 		"ioremap error for 0x%llx-0x%llx, requested 0x%lx, got 0x%lx\n",
 				(unsigned long long)phys_addr,
 				(unsigned long long)(phys_addr + size),
diff --git a/arch/x86/mm/kmemcheck/kmemcheck.c b/arch/x86/mm/kmemcheck/kmemcheck.c
index 528bf95..8cc1833 100644
--- a/arch/x86/mm/kmemcheck/kmemcheck.c
+++ b/arch/x86/mm/kmemcheck/kmemcheck.c
@@ -225,9 +225,6 @@ void kmemcheck_hide(struct pt_regs *regs)
 
 	BUG_ON(!irqs_disabled());
 
-	if (data->balance == 0)
-		return;
-
 	if (unlikely(data->balance != 1)) {
 		kmemcheck_show_all();
 		kmemcheck_error_save_bug(regs);
diff --git a/arch/x86/mm/kmemcheck/shadow.c b/arch/x86/mm/kmemcheck/shadow.c
index e773b6b..3f66b82 100644
--- a/arch/x86/mm/kmemcheck/shadow.c
+++ b/arch/x86/mm/kmemcheck/shadow.c
@@ -1,7 +1,6 @@
 #include <linux/kmemcheck.h>
 #include <linux/module.h>
 #include <linux/mm.h>
-#include <linux/module.h>
 
 #include <asm/page.h>
 #include <asm/pgtable.h>
diff --git a/arch/x86/mm/mmap.c b/arch/x86/mm/mmap.c
index 1658296..c8191de 100644
--- a/arch/x86/mm/mmap.c
+++ b/arch/x86/mm/mmap.c
@@ -29,13 +29,26 @@
 #include <linux/random.h>
 #include <linux/limits.h>
 #include <linux/sched.h>
+#include <asm/elf.h>
+
+static unsigned int stack_maxrandom_size(void)
+{
+	unsigned int max = 0;
+	if ((current->flags & PF_RANDOMIZE) &&
+		!(current->personality & ADDR_NO_RANDOMIZE)) {
+		max = ((-1U) & STACK_RND_MASK) << PAGE_SHIFT;
+	}
+
+	return max;
+}
+
 
 /*
  * Top of mmap area (just below the process stack).
  *
- * Leave an at least ~128 MB hole.
+ * Leave an at least ~128 MB hole with possible stack randomization.
  */
-#define MIN_GAP (128*1024*1024)
+#define MIN_GAP (128*1024*1024UL + stack_maxrandom_size())
 #define MAX_GAP (TASK_SIZE/6*5)
 
 /*
diff --git a/arch/x86/mm/pageattr.c b/arch/x86/mm/pageattr.c
index 7e600c1..dd38bfb 100644
--- a/arch/x86/mm/pageattr.c
+++ b/arch/x86/mm/pageattr.c
@@ -12,6 +12,7 @@
 #include <linux/seq_file.h>
 #include <linux/debugfs.h>
 #include <linux/pfn.h>
+#include <linux/percpu.h>
 
 #include <asm/e820.h>
 #include <asm/processor.h>
@@ -143,6 +144,7 @@ void clflush_cache_range(void *vaddr, unsigned int size)
 
 	mb();
 }
+EXPORT_SYMBOL_GPL(clflush_cache_range);
 
 static void __cpa_flush_all(void *arg)
 {
@@ -686,7 +688,7 @@ static int cpa_process_alias(struct cpa_data *cpa)
 {
 	struct cpa_data alias_cpa;
 	unsigned long laddr = (unsigned long)__va(cpa->pfn << PAGE_SHIFT);
-	unsigned long vaddr, remapped;
+	unsigned long vaddr;
 	int ret;
 
 	if (cpa->pfn >= max_pfn_mapped)
@@ -744,24 +746,6 @@ static int cpa_process_alias(struct cpa_data *cpa)
 	}
 #endif
 
-	/*
-	 * If the PMD page was partially used for per-cpu remapping,
-	 * the recycled area needs to be split and modified.  Because
-	 * the area is always proper subset of a PMD page
-	 * cpa->numpages is guaranteed to be 1 for these areas, so
-	 * there's no need to loop over and check for further remaps.
-	 */
-	remapped = (unsigned long)pcpu_lpage_remapped((void *)laddr);
-	if (remapped) {
-		WARN_ON(cpa->numpages > 1);
-		alias_cpa = *cpa;
-		alias_cpa.vaddr = &remapped;
-		alias_cpa.flags &= ~(CPA_PAGES_ARRAY | CPA_ARRAY);
-		ret = __change_page_attr_set_clr(&alias_cpa, 0);
-		if (ret)
-			return ret;
-	}
-
 	return 0;
 }
 
@@ -822,6 +806,7 @@ static int change_page_attr_set_clr(unsigned long *addr, int numpages,
 {
 	struct cpa_data cpa;
 	int ret, cache, checkalias;
+	unsigned long baddr = 0;
 
 	/*
 	 * Check, if we are requested to change a not supported
@@ -853,6 +838,11 @@ static int change_page_attr_set_clr(unsigned long *addr, int numpages,
 			 */
 			WARN_ON_ONCE(1);
 		}
+		/*
+		 * Save address for cache flush. *addr is modified in the call
+		 * to __change_page_attr_set_clr() below.
+		 */
+		baddr = *addr;
 	}
 
 	/* Must avoid aliasing mappings in the highmem code */
@@ -900,7 +890,7 @@ static int change_page_attr_set_clr(unsigned long *addr, int numpages,
 			cpa_flush_array(addr, numpages, cache,
 					cpa.flags, pages);
 		} else
-			cpa_flush_range(*addr, numpages, cache);
+			cpa_flush_range(baddr, numpages, cache);
 	} else
 		cpa_flush_all(cache);
 
diff --git a/arch/x86/mm/pat.c b/arch/x86/mm/pat.c
index b2f7d3e..e78cd0e 100644
--- a/arch/x86/mm/pat.c
+++ b/arch/x86/mm/pat.c
@@ -15,6 +15,7 @@
 #include <linux/gfp.h>
 #include <linux/mm.h>
 #include <linux/fs.h>
+#include <linux/rbtree.h>
 
 #include <asm/cacheflush.h>
 #include <asm/processor.h>
@@ -80,6 +81,7 @@ enum {
 void pat_init(void)
 {
 	u64 pat;
+	bool boot_cpu = !boot_pat_state;
 
 	if (!pat_enabled)
 		return;
@@ -121,8 +123,10 @@ void pat_init(void)
 		rdmsrl(MSR_IA32_CR_PAT, boot_pat_state);
 
 	wrmsrl(MSR_IA32_CR_PAT, pat);
-	printk(KERN_INFO "x86 PAT enabled: cpu %d, old 0x%Lx, new 0x%Lx\n",
-	       smp_processor_id(), boot_pat_state, pat);
+
+	if (boot_cpu)
+		printk(KERN_INFO "x86 PAT enabled: cpu %d, old 0x%Lx, new 0x%Lx\n",
+		       smp_processor_id(), boot_pat_state, pat);
 }
 
 #undef PAT
@@ -148,11 +152,10 @@ static char *cattr_name(unsigned long flags)
  * areas). All the aliases have the same cache attributes of course.
  * Zero attributes are represented as holes.
  *
- * Currently the data structure is a list because the number of mappings
- * are expected to be relatively small. If this should be a problem
- * it could be changed to a rbtree or similar.
+ * The data structure is a list that is also organized as an rbtree
+ * sorted on the start address of memtype range.
  *
- * memtype_lock protects the whole list.
+ * memtype_lock protects both the linear list and rbtree.
  */
 
 struct memtype {
@@ -160,11 +163,53 @@ struct memtype {
 	u64			end;
 	unsigned long		type;
 	struct list_head	nd;
+	struct rb_node		rb;
 };
 
+static struct rb_root memtype_rbroot = RB_ROOT;
 static LIST_HEAD(memtype_list);
 static DEFINE_SPINLOCK(memtype_lock);	/* protects memtype list */
 
+static struct memtype *memtype_rb_search(struct rb_root *root, u64 start)
+{
+	struct rb_node *node = root->rb_node;
+	struct memtype *last_lower = NULL;
+
+	while (node) {
+		struct memtype *data = container_of(node, struct memtype, rb);
+
+		if (data->start < start) {
+			last_lower = data;
+			node = node->rb_right;
+		} else if (data->start > start) {
+			node = node->rb_left;
+		} else
+			return data;
+	}
+
+	/* Will return NULL if there is no entry with its start <= start */
+	return last_lower;
+}
+
+static void memtype_rb_insert(struct rb_root *root, struct memtype *data)
+{
+	struct rb_node **new = &(root->rb_node);
+	struct rb_node *parent = NULL;
+
+	while (*new) {
+		struct memtype *this = container_of(*new, struct memtype, rb);
+
+		parent = *new;
+		if (data->start <= this->start)
+			new = &((*new)->rb_left);
+		else if (data->start > this->start)
+			new = &((*new)->rb_right);
+	}
+
+	rb_link_node(&data->rb, parent, new);
+	rb_insert_color(&data->rb, root);
+}
+
 /*
  * Does intersection of PAT memory type and MTRR memory type and returns
  * the resulting memory type as PAT understands it.
@@ -218,9 +263,6 @@ chk_conflict(struct memtype *new, struct memtype *entry, unsigned long *type)
 	return -EBUSY;
 }
 
-static struct memtype *cached_entry;
-static u64 cached_start;
-
 static int pat_pagerange_is_ram(unsigned long start, unsigned long end)
 {
 	int ram_page = 0, not_rampage = 0;
@@ -249,63 +291,61 @@ static int pat_pagerange_is_ram(unsigned long start, unsigned long end)
 }
 
 /*
- * For RAM pages, mark the pages as non WB memory type using
- * PageNonWB (PG_arch_1). We allow only one set_memory_uc() or
- * set_memory_wc() on a RAM page at a time before marking it as WB again.
- * This is ok, because only one driver will be owning the page and
- * doing set_memory_*() calls.
+ * For RAM pages, we use page flags to mark the pages with appropriate type.
+ * Here we do two pass:
+ * - Find the memtype of all the pages in the range, look for any conflicts
+ * - In case of no conflicts, set the new memtype for pages in the range
  *
- * For now, we use PageNonWB to track that the RAM page is being mapped
- * as non WB. In future, we will have to use one more flag
- * (or some other mechanism in page_struct) to distinguish between
- * UC and WC mapping.
+ * Caller must hold memtype_lock for atomicity.
  */
 static int reserve_ram_pages_type(u64 start, u64 end, unsigned long req_type,
 				  unsigned long *new_type)
 {
 	struct page *page;
-	u64 pfn, end_pfn;
+	u64 pfn;
+
+	if (req_type == _PAGE_CACHE_UC) {
+		/* We do not support strong UC */
+		WARN_ON_ONCE(1);
+		req_type = _PAGE_CACHE_UC_MINUS;
+	}
 
 	for (pfn = (start >> PAGE_SHIFT); pfn < (end >> PAGE_SHIFT); ++pfn) {
-		page = pfn_to_page(pfn);
-		if (page_mapped(page) || PageNonWB(page))
-			goto out;
+		unsigned long type;
 
-		SetPageNonWB(page);
+		page = pfn_to_page(pfn);
+		type = get_page_memtype(page);
+		if (type != -1) {
+			printk(KERN_INFO "reserve_ram_pages_type failed "
+				"0x%Lx-0x%Lx, track 0x%lx, req 0x%lx\n",
+				start, end, type, req_type);
+			if (new_type)
+				*new_type = type;
+
+			return -EBUSY;
+		}
 	}
-	return 0;
 
-out:
-	end_pfn = pfn;
-	for (pfn = (start >> PAGE_SHIFT); pfn < end_pfn; ++pfn) {
+	if (new_type)
+		*new_type = req_type;
+
+	for (pfn = (start >> PAGE_SHIFT); pfn < (end >> PAGE_SHIFT); ++pfn) {
 		page = pfn_to_page(pfn);
-		ClearPageNonWB(page);
+		set_page_memtype(page, req_type);
 	}
-
-	return -EINVAL;
+	return 0;
 }
 
 static int free_ram_pages_type(u64 start, u64 end)
 {
 	struct page *page;
-	u64 pfn, end_pfn;
+	u64 pfn;
 
 	for (pfn = (start >> PAGE_SHIFT); pfn < (end >> PAGE_SHIFT); ++pfn) {
 		page = pfn_to_page(pfn);
-		if (page_mapped(page) || !PageNonWB(page))
-			goto out;
-
-		ClearPageNonWB(page);
+		set_page_memtype(page, -1);
 	}
 	return 0;
-
-out:
-	end_pfn = pfn;
-	for (pfn = (start >> PAGE_SHIFT); pfn < end_pfn; ++pfn) {
-		page = pfn_to_page(pfn);
-		SetPageNonWB(page);
-	}
-	return -EINVAL;
 }
 
 /*
@@ -339,6 +379,8 @@ int reserve_memtype(u64 start, u64 end, unsigned long req_type,
 		if (new_type) {
 			if (req_type == -1)
 				*new_type = _PAGE_CACHE_WB;
+			else if (req_type == _PAGE_CACHE_WC)
+				*new_type = _PAGE_CACHE_UC_MINUS;
 			else
 				*new_type = req_type & _PAGE_CACHE_MASK;
 		}
@@ -364,11 +406,16 @@ int reserve_memtype(u64 start, u64 end, unsigned long req_type,
 		*new_type = actual_type;
 
 	is_range_ram = pat_pagerange_is_ram(start, end);
-	if (is_range_ram == 1)
-		return reserve_ram_pages_type(start, end, req_type,
-					      new_type);
-	else if (is_range_ram < 0)
+	if (is_range_ram == 1) {
+
+		spin_lock(&memtype_lock);
+		err = reserve_ram_pages_type(start, end, req_type, new_type);
+		spin_unlock(&memtype_lock);
+
+		return err;
+	} else if (is_range_ram < 0) {
 		return -EINVAL;
+	}
 
 	new  = kmalloc(sizeof(struct memtype), GFP_KERNEL);
 	if (!new)
@@ -380,17 +427,11 @@ int reserve_memtype(u64 start, u64 end, unsigned long req_type,
 
 	spin_lock(&memtype_lock);
 
-	if (cached_entry && start >= cached_start)
-		entry = cached_entry;
-	else
-		entry = list_entry(&memtype_list, struct memtype, nd);
-
 	/* Search for existing mapping that overlaps the current range */
 	where = NULL;
-	list_for_each_entry_continue(entry, &memtype_list, nd) {
+	list_for_each_entry(entry, &memtype_list, nd) {
 		if (end <= entry->start) {
 			where = entry->nd.prev;
-			cached_entry = list_entry(where, struct memtype, nd);
 			break;
 		} else if (start <= entry->start) { /* end > entry->start */
 			err = chk_conflict(new, entry, new_type);
@@ -398,8 +439,6 @@ int reserve_memtype(u64 start, u64 end, unsigned long req_type,
 				dprintk("Overlap at 0x%Lx-0x%Lx\n",
 					entry->start, entry->end);
 				where = entry->nd.prev;
-				cached_entry = list_entry(where,
-							struct memtype, nd);
 			}
 			break;
 		} else if (start < entry->end) { /* start > entry->start */
@@ -407,8 +446,6 @@ int reserve_memtype(u64 start, u64 end, unsigned long req_type,
 			if (!err) {
 				dprintk("Overlap at 0x%Lx-0x%Lx\n",
 					entry->start, entry->end);
-				cached_entry = list_entry(entry->nd.prev,
-							struct memtype, nd);
 
 				/*
 				 * Move to right position in the linked
@@ -436,13 +473,13 @@ int reserve_memtype(u64 start, u64 end, unsigned long req_type,
 		return err;
 	}
 
-	cached_start = start;
-
 	if (where)
 		list_add(&new->nd, where);
 	else
 		list_add_tail(&new->nd, &memtype_list);
 
+	memtype_rb_insert(&memtype_rbroot, new);
+
 	spin_unlock(&memtype_lock);
 
 	dprintk("reserve_memtype added 0x%Lx-0x%Lx, track %s, req %s, ret %s\n",
@@ -454,7 +491,7 @@ int reserve_memtype(u64 start, u64 end, unsigned long req_type,
 
 int free_memtype(u64 start, u64 end)
 {
-	struct memtype *entry;
+	struct memtype *entry, *saved_entry;
 	int err = -EINVAL;
 	int is_range_ram;
 
@@ -466,23 +503,58 @@ int free_memtype(u64 start, u64 end)
 		return 0;
 
 	is_range_ram = pat_pagerange_is_ram(start, end);
-	if (is_range_ram == 1)
-		return free_ram_pages_type(start, end);
-	else if (is_range_ram < 0)
+	if (is_range_ram == 1) {
+
+		spin_lock(&memtype_lock);
+		err = free_ram_pages_type(start, end);
+		spin_unlock(&memtype_lock);
+
+		return err;
+	} else if (is_range_ram < 0) {
 		return -EINVAL;
+	}
 
 	spin_lock(&memtype_lock);
-	list_for_each_entry(entry, &memtype_list, nd) {
+
+	entry = memtype_rb_search(&memtype_rbroot, start);
+	if (unlikely(entry == NULL))
+		goto unlock_ret;
+
+	/*
+	 * Saved entry points to an entry with start same or less than what
+	 * we searched for. Now go through the list in both directions to look
+	 * for the entry that matches with both start and end, with list stored
+	 * in sorted start address
+	 */
+	saved_entry = entry;
+	list_for_each_entry_from(entry, &memtype_list, nd) {
 		if (entry->start == start && entry->end == end) {
-			if (cached_entry == entry || cached_start == start)
-				cached_entry = NULL;
+			rb_erase(&entry->rb, &memtype_rbroot);
+			list_del(&entry->nd);
+			kfree(entry);
+			err = 0;
+			break;
+		} else if (entry->start > start) {
+			break;
+		}
+	}
 
+	if (!err)
+		goto unlock_ret;
+
+	entry = saved_entry;
+	list_for_each_entry_reverse(entry, &memtype_list, nd) {
+		if (entry->start == start && entry->end == end) {
+			rb_erase(&entry->rb, &memtype_rbroot);
 			list_del(&entry->nd);
 			kfree(entry);
 			err = 0;
 			break;
+		} else if (entry->start < start) {
+			break;
 		}
 	}
+unlock_ret:
 	spin_unlock(&memtype_lock);
 
 	if (err) {
@@ -496,6 +568,101 @@ int free_memtype(u64 start, u64 end)
 }
 
 
+/**
+ * lookup_memtype - Looksup the memory type for a physical address
+ * @paddr: physical address of which memory type needs to be looked up
+ *
+ * Only to be called when PAT is enabled
+ *
+ * Returns _PAGE_CACHE_WB, _PAGE_CACHE_WC, _PAGE_CACHE_UC_MINUS or
+ * _PAGE_CACHE_UC
+ */
+static unsigned long lookup_memtype(u64 paddr)
+{
+	int rettype = _PAGE_CACHE_WB;
+	struct memtype *entry;
+
+	if (is_ISA_range(paddr, paddr + PAGE_SIZE - 1))
+		return rettype;
+
+	if (pat_pagerange_is_ram(paddr, paddr + PAGE_SIZE)) {
+		struct page *page;
+		spin_lock(&memtype_lock);
+		page = pfn_to_page(paddr >> PAGE_SHIFT);
+		rettype = get_page_memtype(page);
+		spin_unlock(&memtype_lock);
+		/*
+		 * -1 from get_page_memtype() implies RAM page is in its
+		 * default state and not reserved, and hence of type WB
+		 */
+		if (rettype == -1)
+			rettype = _PAGE_CACHE_WB;
+
+		return rettype;
+	}
+
+	spin_lock(&memtype_lock);
+
+	entry = memtype_rb_search(&memtype_rbroot, paddr);
+	if (entry != NULL)
+		rettype = entry->type;
+	else
+		rettype = _PAGE_CACHE_UC_MINUS;
+
+	spin_unlock(&memtype_lock);
+	return rettype;
+}
+
+/**
+ * io_reserve_memtype - Request a memory type mapping for a region of memory
+ * @start: start (physical address) of the region
+ * @end: end (physical address) of the region
+ * @type: A pointer to memtype, with requested type. On success, requested
+ * or any other compatible type that was available for the region is returned
+ *
+ * On success, returns 0
+ * On failure, returns non-zero
+ */
+int io_reserve_memtype(resource_size_t start, resource_size_t end,
+			unsigned long *type)
+{
+	resource_size_t size = end - start;
+	unsigned long req_type = *type;
+	unsigned long new_type;
+	int ret;
+
+	WARN_ON_ONCE(iomem_map_sanity_check(start, size));
+
+	ret = reserve_memtype(start, end, req_type, &new_type);
+	if (ret)
+		goto out_err;
+
+	if (!is_new_memtype_allowed(start, size, req_type, new_type))
+		goto out_free;
+
+	if (kernel_map_sync_memtype(start, size, new_type) < 0)
+		goto out_free;
+
+	*type = new_type;
+	return 0;
+
+out_free:
+	free_memtype(start, end);
+	ret = -EBUSY;
+out_err:
+	return ret;
+}
+
+/**
+ * io_free_memtype - Release a memory type mapping for a region of memory
+ * @start: start (physical address) of the region
+ * @end: end (physical address) of the region
+ */
+void io_free_memtype(resource_size_t start, resource_size_t end)
+{
+	free_memtype(start, end);
+}
+
 pgprot_t phys_mem_access_prot(struct file *file, unsigned long pfn,
 				unsigned long size, pgprot_t vma_prot)
 {
@@ -577,7 +744,7 @@ int kernel_map_sync_memtype(u64 base, unsigned long size, unsigned long flags)
 {
 	unsigned long id_sz;
 
-	if (!pat_enabled || base >= __pa(high_memory))
+	if (base >= __pa(high_memory))
 		return 0;
 
 	id_sz = (__pa(high_memory) < base + size) ?
@@ -612,11 +779,29 @@ static int reserve_pfn_range(u64 paddr, unsigned long size, pgprot_t *vma_prot,
 	is_ram = pat_pagerange_is_ram(paddr, paddr + size);
 
 	/*
-	 * reserve_pfn_range() doesn't support RAM pages. Maintain the current
-	 * behavior with RAM pages by returning success.
+	 * reserve_pfn_range() for RAM pages. We do not refcount to keep
+	 * track of number of mappings of RAM pages. We can assert that
+	 * the type requested matches the type of first page in the range.
 	 */
-	if (is_ram != 0)
+	if (is_ram) {
+		if (!pat_enabled)
+			return 0;
+
+		flags = lookup_memtype(paddr);
+		if (want_flags != flags) {
+			printk(KERN_WARNING
+			"%s:%d map pfn RAM range req %s for %Lx-%Lx, got %s\n",
+				current->comm, current->pid,
+				cattr_name(want_flags),
+				(unsigned long long)paddr,
+				(unsigned long long)(paddr + size),
+				cattr_name(flags));
+			*vma_prot = __pgprot((pgprot_val(*vma_prot) &
+					      (~_PAGE_CACHE_MASK)) |
+					     flags);
+		}
 		return 0;
+	}
 
 	ret = reserve_memtype(paddr, paddr + size, want_flags, &flags);
 	if (ret)
@@ -678,14 +863,6 @@ int track_pfn_vma_copy(struct vm_area_struct *vma)
 	unsigned long vma_size = vma->vm_end - vma->vm_start;
 	pgprot_t pgprot;
 
-	if (!pat_enabled)
-		return 0;
-
-	/*
-	 * For now, only handle remap_pfn_range() vmas where
-	 * is_linear_pfn_mapping() == TRUE. Handling of
-	 * vm_insert_pfn() is TBD.
-	 */
 	if (is_linear_pfn_mapping(vma)) {
 		/*
 		 * reserve the whole chunk covered by vma. We need the
@@ -713,23 +890,24 @@ int track_pfn_vma_copy(struct vm_area_struct *vma)
 int track_pfn_vma_new(struct vm_area_struct *vma, pgprot_t *prot,
 			unsigned long pfn, unsigned long size)
 {
+	unsigned long flags;
 	resource_size_t paddr;
 	unsigned long vma_size = vma->vm_end - vma->vm_start;
 
-	if (!pat_enabled)
-		return 0;
-
-	/*
-	 * For now, only handle remap_pfn_range() vmas where
-	 * is_linear_pfn_mapping() == TRUE. Handling of
-	 * vm_insert_pfn() is TBD.
-	 */
 	if (is_linear_pfn_mapping(vma)) {
 		/* reserve the whole chunk starting from vm_pgoff */
 		paddr = (resource_size_t)vma->vm_pgoff << PAGE_SHIFT;
 		return reserve_pfn_range(paddr, vma_size, prot, 0);
 	}
 
+	if (!pat_enabled)
+		return 0;
+
+	/* for vm_insert_pfn and friends, we set prot based on lookup */
+	flags = lookup_memtype(pfn << PAGE_SHIFT);
+	*prot = __pgprot((pgprot_val(vma->vm_page_prot) & (~_PAGE_CACHE_MASK)) |
+			 flags);
+
 	return 0;
 }
 
@@ -744,14 +922,6 @@ void untrack_pfn_vma(struct vm_area_struct *vma, unsigned long pfn,
 	resource_size_t paddr;
 	unsigned long vma_size = vma->vm_end - vma->vm_start;
 
-	if (!pat_enabled)
-		return;
-
-	/*
-	 * For now, only handle remap_pfn_range() vmas where
-	 * is_linear_pfn_mapping() == TRUE. Handling of
-	 * vm_insert_pfn() is TBD.
-	 */
 	if (is_linear_pfn_mapping(vma)) {
 		/* free the whole chunk starting from vm_pgoff */
 		paddr = (resource_size_t)vma->vm_pgoff << PAGE_SHIFT;
diff --git a/arch/x86/mm/setup_nx.c b/arch/x86/mm/setup_nx.c
new file mode 100644
index 0000000..513d8ed
--- /dev/null
+++ b/arch/x86/mm/setup_nx.c
@@ -0,0 +1,69 @@
+#include <linux/spinlock.h>
+#include <linux/errno.h>
+#include <linux/init.h>
+
+#include <asm/pgtable.h>
+
+int nx_enabled;
+
+#if defined(CONFIG_X86_64) || defined(CONFIG_X86_PAE)
+static int disable_nx __cpuinitdata;
+
+/*
+ * noexec = on|off
+ *
+ * Control non-executable mappings for processes.
+ *
+ * on      Enable
+ * off     Disable
+ */
+static int __init noexec_setup(char *str)
+{
+	if (!str)
+		return -EINVAL;
+	if (!strncmp(str, "on", 2)) {
+		__supported_pte_mask |= _PAGE_NX;
+		disable_nx = 0;
+	} else if (!strncmp(str, "off", 3)) {
+		disable_nx = 1;
+		__supported_pte_mask &= ~_PAGE_NX;
+	}
+	return 0;
+}
+early_param("noexec", noexec_setup);
+#endif
+
+#ifdef CONFIG_X86_PAE
+void __init set_nx(void)
+{
+	unsigned int v[4], l, h;
+
+	if (cpu_has_pae && (cpuid_eax(0x80000000) > 0x80000001)) {
+		cpuid(0x80000001, &v[0], &v[1], &v[2], &v[3]);
+
+		if ((v[3] & (1 << 20)) && !disable_nx) {
+			rdmsr(MSR_EFER, l, h);
+			l |= EFER_NX;
+			wrmsr(MSR_EFER, l, h);
+			nx_enabled = 1;
+			__supported_pte_mask |= _PAGE_NX;
+		}
+	}
+}
+#else
+void set_nx(void)
+{
+}
+#endif
+
+#ifdef CONFIG_X86_64
+void __cpuinit check_efer(void)
+{
+	unsigned long efer;
+
+	rdmsrl(MSR_EFER, efer);
+	if (!(efer & EFER_NX) || disable_nx)
+		__supported_pte_mask &= ~_PAGE_NX;
+}
+#endif
+
diff --git a/arch/x86/mm/testmmiotrace.c b/arch/x86/mm/testmmiotrace.c
index 427fd1b..8565d94 100644
--- a/arch/x86/mm/testmmiotrace.c
+++ b/arch/x86/mm/testmmiotrace.c
@@ -1,12 +1,13 @@
 /*
  * Written by Pekka Paalanen, 2008-2009 <pq@iki.fi>
  */
+
+#define pr_fmt(fmt) KBUILD_MODNAME ": " fmt
+
 #include <linux/module.h>
 #include <linux/io.h>
 #include <linux/mmiotrace.h>
 
-#define MODULE_NAME "testmmiotrace"
-
 static unsigned long mmio_address;
 module_param(mmio_address, ulong, 0);
 MODULE_PARM_DESC(mmio_address, " Start address of the mapping of 16 kB "
@@ -30,7 +31,7 @@ static unsigned v32(unsigned i)
 static void do_write_test(void __iomem *p)
 {
 	unsigned int i;
-	pr_info(MODULE_NAME ": write test.\n");
+	pr_info("write test.\n");
 	mmiotrace_printk("Write test.\n");
 
 	for (i = 0; i < 256; i++)
@@ -47,7 +48,7 @@ static void do_read_test(void __iomem *p)
 {
 	unsigned int i;
 	unsigned errs[3] = { 0 };
-	pr_info(MODULE_NAME ": read test.\n");
+	pr_info("read test.\n");
 	mmiotrace_printk("Read test.\n");
 
 	for (i = 0; i < 256; i++)
@@ -68,7 +69,7 @@ static void do_read_test(void __iomem *p)
 
 static void do_read_far_test(void __iomem *p)
 {
-	pr_info(MODULE_NAME ": read far test.\n");
+	pr_info("read far test.\n");
 	mmiotrace_printk("Read far test.\n");
 
 	ioread32(p + read_far);
@@ -78,7 +79,7 @@ static void do_test(unsigned long size)
 {
 	void __iomem *p = ioremap_nocache(mmio_address, size);
 	if (!p) {
-		pr_err(MODULE_NAME ": could not ioremap, aborting.\n");
+		pr_err("could not ioremap, aborting.\n");
 		return;
 	}
 	mmiotrace_printk("ioremap returned %p.\n", p);
@@ -94,24 +95,22 @@ static int __init init(void)
 	unsigned long size = (read_far) ? (8 << 20) : (16 << 10);
 
 	if (mmio_address == 0) {
-		pr_err(MODULE_NAME ": you have to use the module argument "
-							"mmio_address.\n");
-		pr_err(MODULE_NAME ": DO NOT LOAD THIS MODULE UNLESS"
-				" YOU REALLY KNOW WHAT YOU ARE DOING!\n");
+		pr_err("you have to use the module argument mmio_address.\n");
+		pr_err("DO NOT LOAD THIS MODULE UNLESS YOU REALLY KNOW WHAT YOU ARE DOING!\n");
 		return -ENXIO;
 	}
 
-	pr_warning(MODULE_NAME ": WARNING: mapping %lu kB @ 0x%08lx in PCI "
-		"address space, and writing 16 kB of rubbish in there.\n",
-		 size >> 10, mmio_address);
+	pr_warning("WARNING: mapping %lu kB @ 0x%08lx in PCI address space, "
+		   "and writing 16 kB of rubbish in there.\n",
+		   size >> 10, mmio_address);
 	do_test(size);
-	pr_info(MODULE_NAME ": All done.\n");
+	pr_info("All done.\n");
 	return 0;
 }
 
 static void __exit cleanup(void)
 {
-	pr_debug(MODULE_NAME ": unloaded.\n");
+	pr_debug("unloaded.\n");
 }
 
 module_init(init);
diff --git a/arch/x86/mm/tlb.c b/arch/x86/mm/tlb.c
index c814e14..36fe08e 100644
--- a/arch/x86/mm/tlb.c
+++ b/arch/x86/mm/tlb.c
@@ -59,7 +59,8 @@ void leave_mm(int cpu)
 {
 	if (percpu_read(cpu_tlbstate.state) == TLBSTATE_OK)
 		BUG();
-	cpu_clear(cpu, percpu_read(cpu_tlbstate.active_mm)->cpu_vm_mask);
+	cpumask_clear_cpu(cpu,
+			  mm_cpumask(percpu_read(cpu_tlbstate.active_mm)));
 	load_cr3(swapper_pg_dir);
 }
 EXPORT_SYMBOL_GPL(leave_mm);
@@ -234,8 +235,8 @@ void flush_tlb_current_task(void)
 	preempt_disable();
 
 	local_flush_tlb();
-	if (cpumask_any_but(&mm->cpu_vm_mask, smp_processor_id()) < nr_cpu_ids)
-		flush_tlb_others(&mm->cpu_vm_mask, mm, TLB_FLUSH_ALL);
+	if (cpumask_any_but(mm_cpumask(mm), smp_processor_id()) < nr_cpu_ids)
+		flush_tlb_others(mm_cpumask(mm), mm, TLB_FLUSH_ALL);
 	preempt_enable();
 }
 
@@ -249,8 +250,8 @@ void flush_tlb_mm(struct mm_struct *mm)
 		else
 			leave_mm(smp_processor_id());
 	}
-	if (cpumask_any_but(&mm->cpu_vm_mask, smp_processor_id()) < nr_cpu_ids)
-		flush_tlb_others(&mm->cpu_vm_mask, mm, TLB_FLUSH_ALL);
+	if (cpumask_any_but(mm_cpumask(mm), smp_processor_id()) < nr_cpu_ids)
+		flush_tlb_others(mm_cpumask(mm), mm, TLB_FLUSH_ALL);
 
 	preempt_enable();
 }
@@ -268,8 +269,8 @@ void flush_tlb_page(struct vm_area_struct *vma, unsigned long va)
 			leave_mm(smp_processor_id());
 	}
 
-	if (cpumask_any_but(&mm->cpu_vm_mask, smp_processor_id()) < nr_cpu_ids)
-		flush_tlb_others(&mm->cpu_vm_mask, mm, va);
+	if (cpumask_any_but(mm_cpumask(mm), smp_processor_id()) < nr_cpu_ids)
+		flush_tlb_others(mm_cpumask(mm), mm, va);
 
 	preempt_enable();
 }