[linux-2.6-block.git] / arch / powerpc / mm / pgtable_64.c

/*
 *  This file contains ioremap and related functions for 64-bit machines.
 *
 *  Derived from arch/ppc64/mm/init.c
 *    Copyright (C) 1995-1996 Gary Thomas (gdt@linuxppc.org)
 *
 *  Modifications by Paul Mackerras (PowerMac) (paulus@samba.org)
 *  and Cort Dougan (PReP) (cort@cs.nmt.edu)
 *    Copyright (C) 1996 Paul Mackerras
 *
 *  Derived from "arch/i386/mm/init.c"
 *    Copyright (C) 1991, 1992, 1993, 1994  Linus Torvalds
 *
 *  Dave Engebretsen <engebret@us.ibm.com>
 *      Rework for PPC64 port.
 *
 *  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; either version
 *  2 of the License, or (at your option) any later version.
 *
 */

#include <linux/signal.h>
#include <linux/sched.h>
#include <linux/kernel.h>
#include <linux/errno.h>
#include <linux/string.h>
#include <linux/export.h>
#include <linux/types.h>
#include <linux/mman.h>
#include <linux/mm.h>
#include <linux/swap.h>
#include <linux/stddef.h>
#include <linux/vmalloc.h>
#include <linux/slab.h>
#include <linux/hugetlb.h>

#include <asm/pgalloc.h>
#include <asm/page.h>
#include <asm/prom.h>
#include <asm/io.h>
#include <asm/mmu_context.h>
#include <asm/pgtable.h>
#include <asm/mmu.h>
#include <asm/smp.h>
#include <asm/machdep.h>
#include <asm/tlb.h>
#include <asm/processor.h>
#include <asm/cputable.h>
#include <asm/sections.h>
#include <asm/firmware.h>
#include <asm/dma.h>

#include "mmu_decl.h"


#ifdef CONFIG_PPC_BOOK3S_64
/*
 * partition table and process table for ISA 3.0
 */
struct prtb_entry *process_tb;
struct patb_entry *partition_tb;
/*
 * page table size
 */
unsigned long __pte_index_size;
EXPORT_SYMBOL(__pte_index_size);
unsigned long __pmd_index_size;
EXPORT_SYMBOL(__pmd_index_size);
unsigned long __pud_index_size;
EXPORT_SYMBOL(__pud_index_size);
unsigned long __pgd_index_size;
EXPORT_SYMBOL(__pgd_index_size);
unsigned long __pmd_cache_index;
EXPORT_SYMBOL(__pmd_cache_index);
unsigned long __pud_cache_index;
EXPORT_SYMBOL(__pud_cache_index);
unsigned long __pte_table_size;
EXPORT_SYMBOL(__pte_table_size);
unsigned long __pmd_table_size;
EXPORT_SYMBOL(__pmd_table_size);
unsigned long __pud_table_size;
EXPORT_SYMBOL(__pud_table_size);
unsigned long __pgd_table_size;
EXPORT_SYMBOL(__pgd_table_size);
unsigned long __pmd_val_bits;
EXPORT_SYMBOL(__pmd_val_bits);
unsigned long __pud_val_bits;
EXPORT_SYMBOL(__pud_val_bits);
unsigned long __pgd_val_bits;
EXPORT_SYMBOL(__pgd_val_bits);
unsigned long __kernel_virt_start;
EXPORT_SYMBOL(__kernel_virt_start);
unsigned long __kernel_virt_size;
EXPORT_SYMBOL(__kernel_virt_size);
unsigned long __vmalloc_start;
EXPORT_SYMBOL(__vmalloc_start);
unsigned long __vmalloc_end;
EXPORT_SYMBOL(__vmalloc_end);
unsigned long __kernel_io_start;
EXPORT_SYMBOL(__kernel_io_start);
struct page *vmemmap;
EXPORT_SYMBOL(vmemmap);
unsigned long __pte_frag_nr;
EXPORT_SYMBOL(__pte_frag_nr);
unsigned long __pte_frag_size_shift;
EXPORT_SYMBOL(__pte_frag_size_shift);
unsigned long ioremap_bot;
#else /* !CONFIG_PPC_BOOK3S_64 */
unsigned long ioremap_bot = IOREMAP_BASE;
#endif

/**
 * __ioremap_at - Low level function to establish the page tables
 *                for an IO mapping
 */
void __iomem * __ioremap_at(phys_addr_t pa, void *ea, unsigned long size,
			    unsigned long flags)
{
	unsigned long i;

	/* Make sure we have the base flags */
	if ((flags & _PAGE_PRESENT) == 0)
		flags |= pgprot_val(PAGE_KERNEL);

	/* We don't support the 4K PFN hack with ioremap */
	if (flags & H_PAGE_4K_PFN)
		return NULL;

	WARN_ON(pa & ~PAGE_MASK);
	WARN_ON(((unsigned long)ea) & ~PAGE_MASK);
	WARN_ON(size & ~PAGE_MASK);

	for (i = 0; i < size; i += PAGE_SIZE)
		if (map_kernel_page((unsigned long)ea+i, pa+i, flags))
			return NULL;

	return (void __iomem *)ea;
}

/**
 * __iounmap_from - Low level function to tear down the page tables
 *                  for an IO mapping. This is used for mappings that
 *                  are manipulated manually, like partial unmapping of
 *                  PCI IOs or ISA space.
 */
void __iounmap_at(void *ea, unsigned long size)
{
	WARN_ON(((unsigned long)ea) & ~PAGE_MASK);
	WARN_ON(size & ~PAGE_MASK);

	unmap_kernel_range((unsigned long)ea, size);
}

void __iomem * __ioremap_caller(phys_addr_t addr, unsigned long size,
				unsigned long flags, void *caller)
{
	phys_addr_t paligned;
	void __iomem *ret;

	/*
	 * Choose an address to map it to.
	 * Once the imalloc system is running, we use it.
	 * Before that, we map using addresses going
	 * up from ioremap_bot.  imalloc will use
	 * the addresses from ioremap_bot through
	 * IMALLOC_END
	 * 
	 */
	paligned = addr & PAGE_MASK;
	size = PAGE_ALIGN(addr + size) - paligned;

	if ((size == 0) || (paligned == 0))
		return NULL;

	if (slab_is_available()) {
		struct vm_struct *area;

		area = __get_vm_area_caller(size, VM_IOREMAP,
					    ioremap_bot, IOREMAP_END,
					    caller);
		if (area == NULL)
			return NULL;

		area->phys_addr = paligned;
		ret = __ioremap_at(paligned, area->addr, size, flags);
		if (!ret)
			vunmap(area->addr);
	} else {
		ret = __ioremap_at(paligned, (void *)ioremap_bot, size, flags);
		if (ret)
			ioremap_bot += size;
	}

	if (ret)
		ret += addr & ~PAGE_MASK;
	return ret;
}

void __iomem * __ioremap(phys_addr_t addr, unsigned long size,
			 unsigned long flags)
{
	return __ioremap_caller(addr, size, flags, __builtin_return_address(0));
}

void __iomem * ioremap(phys_addr_t addr, unsigned long size)
{
	unsigned long flags = pgprot_val(pgprot_noncached(__pgprot(0)));
	void *caller = __builtin_return_address(0);

	if (ppc_md.ioremap)
		return ppc_md.ioremap(addr, size, flags, caller);
	return __ioremap_caller(addr, size, flags, caller);
}

void __iomem * ioremap_wc(phys_addr_t addr, unsigned long size)
{
	unsigned long flags = pgprot_val(pgprot_noncached_wc(__pgprot(0)));
	void *caller = __builtin_return_address(0);

	if (ppc_md.ioremap)
		return ppc_md.ioremap(addr, size, flags, caller);
	return __ioremap_caller(addr, size, flags, caller);
}

void __iomem * ioremap_prot(phys_addr_t addr, unsigned long size,
			     unsigned long flags)
{
	void *caller = __builtin_return_address(0);

	/* writeable implies dirty for kernel addresses */
	if (flags & _PAGE_WRITE)
		flags |= _PAGE_DIRTY;

	/* we don't want to let _PAGE_EXEC leak out */
	flags &= ~_PAGE_EXEC;
	/*
	 * Force kernel mapping.
	 */
	flags &= ~_PAGE_USER;
	flags |= _PAGE_PRIVILEGED;

	if (ppc_md.ioremap)
		return ppc_md.ioremap(addr, size, flags, caller);
	return __ioremap_caller(addr, size, flags, caller);
}


/*  
 * Unmap an IO region and remove it from imalloc'd list.
 * Access to IO memory should be serialized by driver.
 */
void __iounmap(volatile void __iomem *token)
{
	void *addr;

	if (!slab_is_available())
		return;
	
	addr = (void *) ((unsigned long __force)
			 PCI_FIX_ADDR(token) & PAGE_MASK);
	if ((unsigned long)addr < ioremap_bot) {
		printk(KERN_WARNING "Attempt to iounmap early bolted mapping"
		       " at 0x%p\n", addr);
		return;
	}
	vunmap(addr);
}

void iounmap(volatile void __iomem *token)
{
	if (ppc_md.iounmap)
		ppc_md.iounmap(token);
	else
		__iounmap(token);
}

EXPORT_SYMBOL(ioremap);
EXPORT_SYMBOL(ioremap_wc);
EXPORT_SYMBOL(ioremap_prot);
EXPORT_SYMBOL(__ioremap);
EXPORT_SYMBOL(__ioremap_at);
EXPORT_SYMBOL(iounmap);
EXPORT_SYMBOL(__iounmap);
EXPORT_SYMBOL(__iounmap_at);

#ifndef __PAGETABLE_PUD_FOLDED
/* 4 level page table */
struct page *pgd_page(pgd_t pgd)
{
	if (pgd_huge(pgd))
		return pte_page(pgd_pte(pgd));
	return virt_to_page(pgd_page_vaddr(pgd));
}
#endif

struct page *pud_page(pud_t pud)
{
	if (pud_huge(pud))
		return pte_page(pud_pte(pud));
	return virt_to_page(pud_page_vaddr(pud));
}

/*
 * For hugepage we have pfn in the pmd, we use PTE_RPN_SHIFT bits for flags
 * For PTE page, we have a PTE_FRAG_SIZE (4K) aligned virtual address.
 */
struct page *pmd_page(pmd_t pmd)
{
	if (pmd_trans_huge(pmd) || pmd_huge(pmd) || pmd_devmap(pmd))
		return pte_page(pmd_pte(pmd));
	return virt_to_page(pmd_page_vaddr(pmd));
}

#ifdef CONFIG_PPC_64K_PAGES
static pte_t *get_from_cache(struct mm_struct *mm)
{
	void *pte_frag, *ret;

	spin_lock(&mm->page_table_lock);
	ret = mm->context.pte_frag;
	if (ret) {
		pte_frag = ret + PTE_FRAG_SIZE;
		/*
		 * If we have taken up all the fragments mark PTE page NULL
		 */
		if (((unsigned long)pte_frag & ~PAGE_MASK) == 0)
			pte_frag = NULL;
		mm->context.pte_frag = pte_frag;
	}
	spin_unlock(&mm->page_table_lock);
	return (pte_t *)ret;
}

static pte_t *__alloc_for_cache(struct mm_struct *mm, int kernel)
{
	void *ret = NULL;
	struct page *page;

	if (!kernel) {
		page = alloc_page(PGALLOC_GFP | __GFP_ACCOUNT);
		if (!page)
			return NULL;
		if (!pgtable_page_ctor(page)) {
			__free_page(page);
			return NULL;
		}
	} else {
		page = alloc_page(PGALLOC_GFP);
		if (!page)
			return NULL;
	}

	ret = page_address(page);
	spin_lock(&mm->page_table_lock);
	/*
	 * If we find pgtable_page set, we return
	 * the allocated page with single fragement
	 * count.
	 */
	if (likely(!mm->context.pte_frag)) {
		set_page_count(page, PTE_FRAG_NR);
		mm->context.pte_frag = ret + PTE_FRAG_SIZE;
	}
	spin_unlock(&mm->page_table_lock);

	return (pte_t *)ret;
}

pte_t *pte_fragment_alloc(struct mm_struct *mm, unsigned long vmaddr, int kernel)
{
	pte_t *pte;

	pte = get_from_cache(mm);
	if (pte)
		return pte;

	return __alloc_for_cache(mm, kernel);
}
#endif /* CONFIG_PPC_64K_PAGES */

void pte_fragment_free(unsigned long *table, int kernel)
{
	struct page *page = virt_to_page(table);
	if (put_page_testzero(page)) {
		if (!kernel)
			pgtable_page_dtor(page);
		free_unref_page(page);
	}
}

#ifdef CONFIG_SMP
void pgtable_free_tlb(struct mmu_gather *tlb, void *table, int shift)
{
	unsigned long pgf = (unsigned long)table;

	BUG_ON(shift > MAX_PGTABLE_INDEX_SIZE);
	pgf |= shift;
	tlb_remove_table(tlb, (void *)pgf);
}

void __tlb_remove_table(void *_table)
{
	void *table = (void *)((unsigned long)_table & ~MAX_PGTABLE_INDEX_SIZE);
	unsigned shift = (unsigned long)_table & MAX_PGTABLE_INDEX_SIZE;

	if (!shift)
		/* PTE page needs special handling */
		pte_fragment_free(table, 0);
	else {
		BUG_ON(shift > MAX_PGTABLE_INDEX_SIZE);
		kmem_cache_free(PGT_CACHE(shift), table);
	}
}
#else
void pgtable_free_tlb(struct mmu_gather *tlb, void *table, int shift)
{
	if (!shift) {
		/* PTE page needs special handling */
		pte_fragment_free(table, 0);
	} else {
		BUG_ON(shift > MAX_PGTABLE_INDEX_SIZE);
		kmem_cache_free(PGT_CACHE(shift), table);
	}
}
#endif

#ifdef CONFIG_STRICT_KERNEL_RWX
void mark_rodata_ro(void)
{
	if (!mmu_has_feature(MMU_FTR_KERNEL_RO)) {
		pr_warn("Warning: Unable to mark rodata read only on this CPU.\n");
		return;
	}

	if (radix_enabled())
		radix__mark_rodata_ro();
	else
		hash__mark_rodata_ro();
}

void mark_initmem_nx(void)
{
	if (radix_enabled())
		radix__mark_initmem_nx();
	else
		hash__mark_initmem_nx();
}
#endif
Commit	Line	Data
14cf11af PM	1	/*
	2	* This file contains ioremap and related functions for 64-bit machines.
	3	*
	4	* Derived from arch/ppc64/mm/init.c
	5	* Copyright (C) 1995-1996 Gary Thomas (gdt@linuxppc.org)
	6	*
	7	* Modifications by Paul Mackerras (PowerMac) (paulus@samba.org)
	8	* and Cort Dougan (PReP) (cort@cs.nmt.edu)
	9	* Copyright (C) 1996 Paul Mackerras
14cf11af PM	10	*
	11	* Derived from "arch/i386/mm/init.c"
	12	* Copyright (C) 1991, 1992, 1993, 1994 Linus Torvalds
	13	*
	14	* Dave Engebretsen <engebret@us.ibm.com>
	15	* Rework for PPC64 port.
	16	*
	17	* This program is free software; you can redistribute it and/or
	18	* modify it under the terms of the GNU General Public License
	19	* as published by the Free Software Foundation; either version
	20	* 2 of the License, or (at your option) any later version.
	21	*
	22	*/
	23
14cf11af PM	24	#include <linux/signal.h>
	25	#include <linux/sched.h>
	26	#include <linux/kernel.h>
	27	#include <linux/errno.h>
	28	#include <linux/string.h>
66b15db6	29	#include <linux/export.h>
14cf11af PM	30	#include <linux/types.h>
	31	#include <linux/mman.h>
	32	#include <linux/mm.h>
	33	#include <linux/swap.h>
	34	#include <linux/stddef.h>
	35	#include <linux/vmalloc.h>
5a0e3ad6	36	#include <linux/slab.h>
06743521	37	#include <linux/hugetlb.h>
14cf11af PM	38
	39	#include <asm/pgalloc.h>
	40	#include <asm/page.h>
	41	#include <asm/prom.h>
14cf11af PM	42	#include <asm/io.h>
	43	#include <asm/mmu_context.h>
	44	#include <asm/pgtable.h>
	45	#include <asm/mmu.h>
14cf11af PM	46	#include <asm/smp.h>
	47	#include <asm/machdep.h>
	48	#include <asm/tlb.h>
14cf11af	49	#include <asm/processor.h>
14cf11af	50	#include <asm/cputable.h>
14cf11af	51	#include <asm/sections.h>
5e203d68	52	#include <asm/firmware.h>
68cf0d64	53	#include <asm/dma.h>
800fc3ee DG	54
800fc3ee DG	55	#include "mmu_decl.h"
14cf11af	56
14cf11af	57
50de596d AK	58	#ifdef CONFIG_PPC_BOOK3S_64
	59	/*
	60	* partition table and process table for ISA 3.0
	61	*/
	62	struct prtb_entry *process_tb;
	63	struct patb_entry *partition_tb;
dd1842a2 AK	64	/*
	65	* page table size
	66	*/
	67	unsigned long __pte_index_size;
	68	EXPORT_SYMBOL(__pte_index_size);
	69	unsigned long __pmd_index_size;
	70	EXPORT_SYMBOL(__pmd_index_size);
	71	unsigned long __pud_index_size;
	72	EXPORT_SYMBOL(__pud_index_size);
	73	unsigned long __pgd_index_size;
	74	EXPORT_SYMBOL(__pgd_index_size);
	75	unsigned long __pmd_cache_index;
	76	EXPORT_SYMBOL(__pmd_cache_index);
fae22116 AK	77	unsigned long __pud_cache_index;
fae22116 AK	78	EXPORT_SYMBOL(__pud_cache_index);
dd1842a2 AK	79	unsigned long __pte_table_size;
	80	EXPORT_SYMBOL(__pte_table_size);
	81	unsigned long __pmd_table_size;
	82	EXPORT_SYMBOL(__pmd_table_size);
	83	unsigned long __pud_table_size;
	84	EXPORT_SYMBOL(__pud_table_size);
	85	unsigned long __pgd_table_size;
	86	EXPORT_SYMBOL(__pgd_table_size);
a2f41eb9 AK	87	unsigned long __pmd_val_bits;
	88	EXPORT_SYMBOL(__pmd_val_bits);
	89	unsigned long __pud_val_bits;
	90	EXPORT_SYMBOL(__pud_val_bits);
	91	unsigned long __pgd_val_bits;
	92	EXPORT_SYMBOL(__pgd_val_bits);
d6a9996e AK	93	unsigned long __kernel_virt_start;
	94	EXPORT_SYMBOL(__kernel_virt_start);
	95	unsigned long __kernel_virt_size;
	96	EXPORT_SYMBOL(__kernel_virt_size);
	97	unsigned long __vmalloc_start;
	98	EXPORT_SYMBOL(__vmalloc_start);
	99	unsigned long __vmalloc_end;
	100	EXPORT_SYMBOL(__vmalloc_end);
63ee9b2f ME	101	unsigned long __kernel_io_start;
63ee9b2f ME	102	EXPORT_SYMBOL(__kernel_io_start);
d6a9996e AK	103	struct page *vmemmap;
d6a9996e AK	104	EXPORT_SYMBOL(vmemmap);
5ed7ecd0 AK	105	unsigned long __pte_frag_nr;
	106	EXPORT_SYMBOL(__pte_frag_nr);
	107	unsigned long __pte_frag_size_shift;
	108	EXPORT_SYMBOL(__pte_frag_size_shift);
d6a9996e AK	109	unsigned long ioremap_bot;
d6a9996e AK	110	#else /* !CONFIG_PPC_BOOK3S_64 */
78f1dbde	111	unsigned long ioremap_bot = IOREMAP_BASE;
d6a9996e	112	#endif
a245067e	113
3d5134ee BH	114	/**
	115	* __ioremap_at - Low level function to establish the page tables
	116	* for an IO mapping
	117	*/
	118	void __iomem * __ioremap_at(phys_addr_t pa, void *ea, unsigned long size,
14cf11af PM	119	unsigned long flags)
	120	{
	121	unsigned long i;
	122
a1f242ff	123	/* Make sure we have the base flags */
14cf11af PM	124	if ((flags & _PAGE_PRESENT) == 0)
	125	flags \|= pgprot_val(PAGE_KERNEL);
	126
a1f242ff	127	/* We don't support the 4K PFN hack with ioremap */
945537df	128	if (flags & H_PAGE_4K_PFN)
a1f242ff BH	129	return NULL;
a1f242ff BH	130
3d5134ee BH	131	WARN_ON(pa & ~PAGE_MASK);
	132	WARN_ON(((unsigned long)ea) & ~PAGE_MASK);
	133	WARN_ON(size & ~PAGE_MASK);
	134
14cf11af	135	for (i = 0; i < size; i += PAGE_SIZE)
a245067e	136	if (map_kernel_page((unsigned long)ea+i, pa+i, flags))
14cf11af PM	137	return NULL;
14cf11af PM	138
3d5134ee BH	139	return (void __iomem *)ea;
	140	}
	141
	142	/**
	143	* __iounmap_from - Low level function to tear down the page tables
	144	* for an IO mapping. This is used for mappings that
	145	* are manipulated manually, like partial unmapping of
	146	* PCI IOs or ISA space.
	147	*/
	148	void __iounmap_at(void *ea, unsigned long size)
	149	{
	150	WARN_ON(((unsigned long)ea) & ~PAGE_MASK);
	151	WARN_ON(size & ~PAGE_MASK);
	152
	153	unmap_kernel_range((unsigned long)ea, size);
14cf11af PM	154	}
14cf11af PM	155
1cdab55d BH	156	void __iomem * __ioremap_caller(phys_addr_t addr, unsigned long size,
1cdab55d BH	157	unsigned long flags, void *caller)
14cf11af	158	{
3d5134ee	159	phys_addr_t paligned;
14cf11af PM	160	void __iomem *ret;
	161
	162	/*
	163	* Choose an address to map it to.
	164	* Once the imalloc system is running, we use it.
	165	* Before that, we map using addresses going
	166	* up from ioremap_bot. imalloc will use
	167	* the addresses from ioremap_bot through
	168	* IMALLOC_END
	169	*
	170	*/
3d5134ee BH	171	paligned = addr & PAGE_MASK;
3d5134ee BH	172	size = PAGE_ALIGN(addr + size) - paligned;
14cf11af	173
3d5134ee	174	if ((size == 0) \|\| (paligned == 0))
14cf11af PM	175	return NULL;
14cf11af PM	176
f691fa10	177	if (slab_is_available()) {
14cf11af	178	struct vm_struct *area;
3d5134ee	179
1cdab55d BH	180	area = __get_vm_area_caller(size, VM_IOREMAP,
	181	ioremap_bot, IOREMAP_END,
	182	caller);
14cf11af PM	183	if (area == NULL)
14cf11af PM	184	return NULL;
7a9d1256 ME	185
7a9d1256 ME	186	area->phys_addr = paligned;
3d5134ee	187	ret = __ioremap_at(paligned, area->addr, size, flags);
14cf11af	188	if (!ret)
3d5134ee	189	vunmap(area->addr);
14cf11af	190	} else {
3d5134ee	191	ret = __ioremap_at(paligned, (void *)ioremap_bot, size, flags);
14cf11af PM	192	if (ret)
	193	ioremap_bot += size;
	194	}
3d5134ee BH	195
	196	if (ret)
	197	ret += addr & ~PAGE_MASK;
14cf11af PM	198	return ret;
	199	}
	200
1cdab55d BH	201	void __iomem * __ioremap(phys_addr_t addr, unsigned long size,
	202	unsigned long flags)
	203	{
	204	return __ioremap_caller(addr, size, flags, __builtin_return_address(0));
	205	}
4cb3cee0	206
68a64357	207	void __iomem * ioremap(phys_addr_t addr, unsigned long size)
4cb3cee0	208	{
72176dd0	209	unsigned long flags = pgprot_val(pgprot_noncached(__pgprot(0)));
1cdab55d	210	void *caller = __builtin_return_address(0);
4cb3cee0 BH	211
4cb3cee0 BH	212	if (ppc_md.ioremap)
1cdab55d BH	213	return ppc_md.ioremap(addr, size, flags, caller);
1cdab55d BH	214	return __ioremap_caller(addr, size, flags, caller);
4cb3cee0 BH	215	}
4cb3cee0 BH	216
be135f40 AB	217	void __iomem * ioremap_wc(phys_addr_t addr, unsigned long size)
be135f40 AB	218	{
72176dd0	219	unsigned long flags = pgprot_val(pgprot_noncached_wc(__pgprot(0)));
be135f40 AB	220	void *caller = __builtin_return_address(0);
	221
	222	if (ppc_md.ioremap)
	223	return ppc_md.ioremap(addr, size, flags, caller);
	224	return __ioremap_caller(addr, size, flags, caller);
	225	}
	226
40f1ce7f	227	void __iomem * ioremap_prot(phys_addr_t addr, unsigned long size,
4cb3cee0 BH	228	unsigned long flags)
4cb3cee0 BH	229	{
1cdab55d BH	230	void *caller = __builtin_return_address(0);
1cdab55d BH	231
a1f242ff	232	/* writeable implies dirty for kernel addresses */
c7d54842	233	if (flags & _PAGE_WRITE)
a1f242ff BH	234	flags \|= _PAGE_DIRTY;
a1f242ff BH	235
ac29c640 AK	236	/* we don't want to let _PAGE_EXEC leak out */
	237	flags &= ~_PAGE_EXEC;
	238	/*
	239	* Force kernel mapping.
	240	*/
ac29c640	241	flags &= ~_PAGE_USER;
812fadcb	242	flags \|= _PAGE_PRIVILEGED;
55052eec	243
4cb3cee0	244	if (ppc_md.ioremap)
1cdab55d BH	245	return ppc_md.ioremap(addr, size, flags, caller);
1cdab55d BH	246	return __ioremap_caller(addr, size, flags, caller);
4cb3cee0 BH	247	}
	248
	249
14cf11af PM	250	/*
	251	* Unmap an IO region and remove it from imalloc'd list.
	252	* Access to IO memory should be serialized by driver.
14cf11af	253	*/
68a64357	254	void __iounmap(volatile void __iomem *token)
14cf11af PM	255	{
	256	void *addr;
	257
f691fa10	258	if (!slab_is_available())
14cf11af PM	259	return;
14cf11af PM	260
3d5134ee BH	261	addr = (void *) ((unsigned long __force)
	262	PCI_FIX_ADDR(token) & PAGE_MASK);
	263	if ((unsigned long)addr < ioremap_bot) {
	264	printk(KERN_WARNING "Attempt to iounmap early bolted mapping"
	265	" at 0x%p\n", addr);
	266	return;
	267	}
	268	vunmap(addr);
14cf11af PM	269	}
14cf11af PM	270
68a64357	271	void iounmap(volatile void __iomem *token)
4cb3cee0 BH	272	{
	273	if (ppc_md.iounmap)
	274	ppc_md.iounmap(token);
	275	else
	276	__iounmap(token);
	277	}
	278
14cf11af	279	EXPORT_SYMBOL(ioremap);
be135f40	280	EXPORT_SYMBOL(ioremap_wc);
40f1ce7f	281	EXPORT_SYMBOL(ioremap_prot);
14cf11af	282	EXPORT_SYMBOL(__ioremap);
a302cb9d	283	EXPORT_SYMBOL(__ioremap_at);
14cf11af	284	EXPORT_SYMBOL(iounmap);
4cb3cee0	285	EXPORT_SYMBOL(__iounmap);
a302cb9d	286	EXPORT_SYMBOL(__iounmap_at);
5c1f6ee9	287
06743521 AK	288	#ifndef __PAGETABLE_PUD_FOLDED
	289	/* 4 level page table */
	290	struct page *pgd_page(pgd_t pgd)
	291	{
	292	if (pgd_huge(pgd))
	293	return pte_page(pgd_pte(pgd));
	294	return virt_to_page(pgd_page_vaddr(pgd));
	295	}
	296	#endif
	297
	298	struct page *pud_page(pud_t pud)
	299	{
	300	if (pud_huge(pud))
	301	return pte_page(pud_pte(pud));
	302	return virt_to_page(pud_page_vaddr(pud));
	303	}
	304
074c2eae AK	305	/*
	306	* For hugepage we have pfn in the pmd, we use PTE_RPN_SHIFT bits for flags
	307	* For PTE page, we have a PTE_FRAG_SIZE (4K) aligned virtual address.
	308	*/
	309	struct page *pmd_page(pmd_t pmd)
	310	{
ebd31197	311	if (pmd_trans_huge(pmd) \|\| pmd_huge(pmd) \|\| pmd_devmap(pmd))
e34aa03c	312	return pte_page(pmd_pte(pmd));
074c2eae AK	313	return virt_to_page(pmd_page_vaddr(pmd));
	314	}
	315
5c1f6ee9 AK	316	#ifdef CONFIG_PPC_64K_PAGES
	317	static pte_t get_from_cache(struct mm_struct mm)
	318	{
	319	void pte_frag, ret;
	320
	321	spin_lock(&mm->page_table_lock);
	322	ret = mm->context.pte_frag;
	323	if (ret) {
	324	pte_frag = ret + PTE_FRAG_SIZE;
	325	/*
	326	* If we have taken up all the fragments mark PTE page NULL
	327	*/
	328	if (((unsigned long)pte_frag & ~PAGE_MASK) == 0)
	329	pte_frag = NULL;
	330	mm->context.pte_frag = pte_frag;
	331	}
	332	spin_unlock(&mm->page_table_lock);
	333	return (pte_t *)ret;
	334	}
	335
	336	static pte_t __alloc_for_cache(struct mm_struct mm, int kernel)
	337	{
	338	void *ret = NULL;
de3b8761 BS	339	struct page *page;
	340
	341	if (!kernel) {
	342	page = alloc_page(PGALLOC_GFP \| __GFP_ACCOUNT);
	343	if (!page)
	344	return NULL;
	345	if (!pgtable_page_ctor(page)) {
	346	__free_page(page);
	347	return NULL;
	348	}
	349	} else {
	350	page = alloc_page(PGALLOC_GFP);
	351	if (!page)
	352	return NULL;
4f804943	353	}
5c1f6ee9 AK	354
	355	ret = page_address(page);
	356	spin_lock(&mm->page_table_lock);
	357	/*
	358	* If we find pgtable_page set, we return
	359	* the allocated page with single fragement
	360	* count.
	361	*/
	362	if (likely(!mm->context.pte_frag)) {
fe896d18	363	set_page_count(page, PTE_FRAG_NR);
5c1f6ee9 AK	364	mm->context.pte_frag = ret + PTE_FRAG_SIZE;
	365	}
	366	spin_unlock(&mm->page_table_lock);
	367
5c1f6ee9 AK	368	return (pte_t *)ret;
	369	}
	370
74701d59	371	pte_t pte_fragment_alloc(struct mm_struct mm, unsigned long vmaddr, int kernel)
5c1f6ee9 AK	372	{
	373	pte_t *pte;
	374
	375	pte = get_from_cache(mm);
	376	if (pte)
	377	return pte;
	378
	379	return __alloc_for_cache(mm, kernel);
	380	}
934828ed	381	#endif /* CONFIG_PPC_64K_PAGES */
5c1f6ee9	382
74701d59	383	void pte_fragment_free(unsigned long *table, int kernel)
5c1f6ee9 AK	384	{
	385	struct page *page = virt_to_page(table);
	386	if (put_page_testzero(page)) {
	387	if (!kernel)
	388	pgtable_page_dtor(page);
2d4894b5	389	free_unref_page(page);
5c1f6ee9 AK	390	}
	391	}
	392
	393	#ifdef CONFIG_SMP
5c1f6ee9 AK	394	void pgtable_free_tlb(struct mmu_gather tlb, void table, int shift)
	395	{
	396	unsigned long pgf = (unsigned long)table;
	397
	398	BUG_ON(shift > MAX_PGTABLE_INDEX_SIZE);
	399	pgf \|= shift;
	400	tlb_remove_table(tlb, (void *)pgf);
	401	}
	402
	403	void __tlb_remove_table(void *_table)
	404	{
	405	void table = (void )((unsigned long)_table & ~MAX_PGTABLE_INDEX_SIZE);
	406	unsigned shift = (unsigned long)_table & MAX_PGTABLE_INDEX_SIZE;
	407
	408	if (!shift)
	409	/* PTE page needs special handling */
74701d59	410	pte_fragment_free(table, 0);
5c1f6ee9 AK	411	else {
	412	BUG_ON(shift > MAX_PGTABLE_INDEX_SIZE);
	413	kmem_cache_free(PGT_CACHE(shift), table);
	414	}
	415	}
	416	#else
	417	void pgtable_free_tlb(struct mmu_gather tlb, void table, int shift)
	418	{
	419	if (!shift) {
	420	/* PTE page needs special handling */
74701d59	421	pte_fragment_free(table, 0);
5c1f6ee9 AK	422	} else {
	423	BUG_ON(shift > MAX_PGTABLE_INDEX_SIZE);
	424	kmem_cache_free(PGT_CACHE(shift), table);
	425	}
	426	}
	427	#endif
9d661958	428
cd65d697 BS	429	#ifdef CONFIG_STRICT_KERNEL_RWX
	430	void mark_rodata_ro(void)
	431	{
	432	if (!mmu_has_feature(MMU_FTR_KERNEL_RO)) {
	433	pr_warn("Warning: Unable to mark rodata read only on this CPU.\n");
	434	return;
	435	}
	436
7614ff32 BS	437	if (radix_enabled())
	438	radix__mark_rodata_ro();
	439	else
cd65d697 BS	440	hash__mark_rodata_ro();
cd65d697 BS	441	}
029d9252 ME	442
	443	void mark_initmem_nx(void)
	444	{
	445	if (radix_enabled())
	446	radix__mark_initmem_nx();
	447	else
	448	hash__mark_initmem_nx();
	449	}
cd65d697	450	#endif