[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/memblock.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/trace.h>
#include <asm/processor.h>
#include <asm/cputable.h>
#include <asm/sections.h>
#include <asm/firmware.h>
#include <asm/dma.h>
#include <asm/powernv.h>

#include "mmu_decl.h"

#ifdef CONFIG_PPC_BOOK3S_64
#if TASK_SIZE_USER64 > (1UL << (ESID_BITS + SID_SHIFT))
#error TASK_SIZE_USER64 exceeds user VSID range
#endif
#endif

#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 __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_PPC_BOOK3S_64
void __init mmu_partition_table_init(void)
{
	unsigned long patb_size = 1UL << PATB_SIZE_SHIFT;
	unsigned long ptcr;

	BUILD_BUG_ON_MSG((PATB_SIZE_SHIFT > 36), "Partition table size too large.");
	partition_tb = __va(memblock_alloc_base(patb_size, patb_size,
						MEMBLOCK_ALLOC_ANYWHERE));

	/* Initialize the Partition Table with no entries */
	memset((void *)partition_tb, 0, patb_size);

	/*
	 * update partition table control register,
	 * 64 K size.
	 */
	ptcr = __pa(partition_tb) | (PATB_SIZE_SHIFT - 12);
	mtspr(SPRN_PTCR, ptcr);
	powernv_set_nmmu_ptcr(ptcr);
}

void mmu_partition_table_set_entry(unsigned int lpid, unsigned long dw0,
				   unsigned long dw1)
{
	unsigned long old = be64_to_cpu(partition_tb[lpid].patb0);

	partition_tb[lpid].patb0 = cpu_to_be64(dw0);
	partition_tb[lpid].patb1 = cpu_to_be64(dw1);

	/*
	 * Global flush of TLBs and partition table caches for this lpid.
	 * The type of flush (hash or radix) depends on what the previous
	 * use of this partition ID was, not the new use.
	 */
	asm volatile("ptesync" : : : "memory");
	if (old & PATB_HR) {
		asm volatile(PPC_TLBIE_5(%0,%1,2,0,1) : :
			     "r" (TLBIEL_INVAL_SET_LPID), "r" (lpid));
		asm volatile(PPC_TLBIE_5(%0,%1,2,1,1) : :
			     "r" (TLBIEL_INVAL_SET_LPID), "r" (lpid));
		trace_tlbie(lpid, 0, TLBIEL_INVAL_SET_LPID, lpid, 2, 0, 1);
	} else {
		asm volatile(PPC_TLBIE_5(%0,%1,2,0,0) : :
			     "r" (TLBIEL_INVAL_SET_LPID), "r" (lpid));
		trace_tlbie(lpid, 0, TLBIEL_INVAL_SET_LPID, lpid, 2, 0, 0);
	}
	asm volatile("eieio; tlbsync; ptesync" : : : "memory");
}
EXPORT_SYMBOL_GPL(mmu_partition_table_set_entry);
#endif /* CONFIG_PPC_BOOK3S_64 */

#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>
95f72d1e	36	#include <linux/memblock.h>
5a0e3ad6	37	#include <linux/slab.h>
06743521	38	#include <linux/hugetlb.h>
14cf11af PM	39
	40	#include <asm/pgalloc.h>
	41	#include <asm/page.h>
	42	#include <asm/prom.h>
14cf11af PM	43	#include <asm/io.h>
	44	#include <asm/mmu_context.h>
	45	#include <asm/pgtable.h>
	46	#include <asm/mmu.h>
14cf11af PM	47	#include <asm/smp.h>
	48	#include <asm/machdep.h>
	49	#include <asm/tlb.h>
0428491c	50	#include <asm/trace.h>
14cf11af	51	#include <asm/processor.h>
14cf11af	52	#include <asm/cputable.h>
14cf11af	53	#include <asm/sections.h>
5e203d68	54	#include <asm/firmware.h>
68cf0d64	55	#include <asm/dma.h>
1d0761d2	56	#include <asm/powernv.h>
800fc3ee DG	57
800fc3ee DG	58	#include "mmu_decl.h"
14cf11af	59
4e003747	60	#ifdef CONFIG_PPC_BOOK3S_64
af81d787	61	#if TASK_SIZE_USER64 > (1UL << (ESID_BITS + SID_SHIFT))
78f1dbde AK	62	#error TASK_SIZE_USER64 exceeds user VSID range
	63	#endif
	64	#endif
14cf11af	65
50de596d AK	66	#ifdef CONFIG_PPC_BOOK3S_64
	67	/*
	68	* partition table and process table for ISA 3.0
	69	*/
	70	struct prtb_entry *process_tb;
	71	struct patb_entry *partition_tb;
dd1842a2 AK	72	/*
	73	* page table size
	74	*/
	75	unsigned long __pte_index_size;
	76	EXPORT_SYMBOL(__pte_index_size);
	77	unsigned long __pmd_index_size;
	78	EXPORT_SYMBOL(__pmd_index_size);
	79	unsigned long __pud_index_size;
	80	EXPORT_SYMBOL(__pud_index_size);
	81	unsigned long __pgd_index_size;
	82	EXPORT_SYMBOL(__pgd_index_size);
	83	unsigned long __pmd_cache_index;
	84	EXPORT_SYMBOL(__pmd_cache_index);
	85	unsigned long __pte_table_size;
	86	EXPORT_SYMBOL(__pte_table_size);
	87	unsigned long __pmd_table_size;
	88	EXPORT_SYMBOL(__pmd_table_size);
	89	unsigned long __pud_table_size;
	90	EXPORT_SYMBOL(__pud_table_size);
	91	unsigned long __pgd_table_size;
	92	EXPORT_SYMBOL(__pgd_table_size);
a2f41eb9 AK	93	unsigned long __pmd_val_bits;
	94	EXPORT_SYMBOL(__pmd_val_bits);
	95	unsigned long __pud_val_bits;
	96	EXPORT_SYMBOL(__pud_val_bits);
	97	unsigned long __pgd_val_bits;
	98	EXPORT_SYMBOL(__pgd_val_bits);
d6a9996e AK	99	unsigned long __kernel_virt_start;
	100	EXPORT_SYMBOL(__kernel_virt_start);
	101	unsigned long __kernel_virt_size;
	102	EXPORT_SYMBOL(__kernel_virt_size);
	103	unsigned long __vmalloc_start;
	104	EXPORT_SYMBOL(__vmalloc_start);
	105	unsigned long __vmalloc_end;
	106	EXPORT_SYMBOL(__vmalloc_end);
63ee9b2f ME	107	unsigned long __kernel_io_start;
63ee9b2f ME	108	EXPORT_SYMBOL(__kernel_io_start);
d6a9996e AK	109	struct page *vmemmap;
d6a9996e AK	110	EXPORT_SYMBOL(vmemmap);
5ed7ecd0 AK	111	unsigned long __pte_frag_nr;
	112	EXPORT_SYMBOL(__pte_frag_nr);
	113	unsigned long __pte_frag_size_shift;
	114	EXPORT_SYMBOL(__pte_frag_size_shift);
d6a9996e AK	115	unsigned long ioremap_bot;
d6a9996e AK	116	#else /* !CONFIG_PPC_BOOK3S_64 */
78f1dbde	117	unsigned long ioremap_bot = IOREMAP_BASE;
d6a9996e	118	#endif
a245067e	119
3d5134ee BH	120	/**
	121	* __ioremap_at - Low level function to establish the page tables
	122	* for an IO mapping
	123	*/
	124	void __iomem * __ioremap_at(phys_addr_t pa, void *ea, unsigned long size,
14cf11af PM	125	unsigned long flags)
	126	{
	127	unsigned long i;
	128
a1f242ff	129	/* Make sure we have the base flags */
14cf11af PM	130	if ((flags & _PAGE_PRESENT) == 0)
	131	flags \|= pgprot_val(PAGE_KERNEL);
	132
a1f242ff	133	/* We don't support the 4K PFN hack with ioremap */
945537df	134	if (flags & H_PAGE_4K_PFN)
a1f242ff BH	135	return NULL;
a1f242ff BH	136
3d5134ee BH	137	WARN_ON(pa & ~PAGE_MASK);
	138	WARN_ON(((unsigned long)ea) & ~PAGE_MASK);
	139	WARN_ON(size & ~PAGE_MASK);
	140
14cf11af	141	for (i = 0; i < size; i += PAGE_SIZE)
a245067e	142	if (map_kernel_page((unsigned long)ea+i, pa+i, flags))
14cf11af PM	143	return NULL;
14cf11af PM	144
3d5134ee BH	145	return (void __iomem *)ea;
	146	}
	147
	148	/**
	149	* __iounmap_from - Low level function to tear down the page tables
	150	* for an IO mapping. This is used for mappings that
	151	* are manipulated manually, like partial unmapping of
	152	* PCI IOs or ISA space.
	153	*/
	154	void __iounmap_at(void *ea, unsigned long size)
	155	{
	156	WARN_ON(((unsigned long)ea) & ~PAGE_MASK);
	157	WARN_ON(size & ~PAGE_MASK);
	158
	159	unmap_kernel_range((unsigned long)ea, size);
14cf11af PM	160	}
14cf11af PM	161
1cdab55d BH	162	void __iomem * __ioremap_caller(phys_addr_t addr, unsigned long size,
1cdab55d BH	163	unsigned long flags, void *caller)
14cf11af	164	{
3d5134ee	165	phys_addr_t paligned;
14cf11af PM	166	void __iomem *ret;
	167
	168	/*
	169	* Choose an address to map it to.
	170	* Once the imalloc system is running, we use it.
	171	* Before that, we map using addresses going
	172	* up from ioremap_bot. imalloc will use
	173	* the addresses from ioremap_bot through
	174	* IMALLOC_END
	175	*
	176	*/
3d5134ee BH	177	paligned = addr & PAGE_MASK;
3d5134ee BH	178	size = PAGE_ALIGN(addr + size) - paligned;
14cf11af	179
3d5134ee	180	if ((size == 0) \|\| (paligned == 0))
14cf11af PM	181	return NULL;
14cf11af PM	182
f691fa10	183	if (slab_is_available()) {
14cf11af	184	struct vm_struct *area;
3d5134ee	185
1cdab55d BH	186	area = __get_vm_area_caller(size, VM_IOREMAP,
	187	ioremap_bot, IOREMAP_END,
	188	caller);
14cf11af PM	189	if (area == NULL)
14cf11af PM	190	return NULL;
7a9d1256 ME	191
7a9d1256 ME	192	area->phys_addr = paligned;
3d5134ee	193	ret = __ioremap_at(paligned, area->addr, size, flags);
14cf11af	194	if (!ret)
3d5134ee	195	vunmap(area->addr);
14cf11af	196	} else {
3d5134ee	197	ret = __ioremap_at(paligned, (void *)ioremap_bot, size, flags);
14cf11af PM	198	if (ret)
	199	ioremap_bot += size;
	200	}
3d5134ee BH	201
	202	if (ret)
	203	ret += addr & ~PAGE_MASK;
14cf11af PM	204	return ret;
	205	}
	206
1cdab55d BH	207	void __iomem * __ioremap(phys_addr_t addr, unsigned long size,
	208	unsigned long flags)
	209	{
	210	return __ioremap_caller(addr, size, flags, __builtin_return_address(0));
	211	}
4cb3cee0	212
68a64357	213	void __iomem * ioremap(phys_addr_t addr, unsigned long size)
4cb3cee0	214	{
72176dd0	215	unsigned long flags = pgprot_val(pgprot_noncached(__pgprot(0)));
1cdab55d	216	void *caller = __builtin_return_address(0);
4cb3cee0 BH	217
4cb3cee0 BH	218	if (ppc_md.ioremap)
1cdab55d BH	219	return ppc_md.ioremap(addr, size, flags, caller);
1cdab55d BH	220	return __ioremap_caller(addr, size, flags, caller);
4cb3cee0 BH	221	}
4cb3cee0 BH	222
be135f40 AB	223	void __iomem * ioremap_wc(phys_addr_t addr, unsigned long size)
be135f40 AB	224	{
72176dd0	225	unsigned long flags = pgprot_val(pgprot_noncached_wc(__pgprot(0)));
be135f40 AB	226	void *caller = __builtin_return_address(0);
	227
	228	if (ppc_md.ioremap)
	229	return ppc_md.ioremap(addr, size, flags, caller);
	230	return __ioremap_caller(addr, size, flags, caller);
	231	}
	232
40f1ce7f	233	void __iomem * ioremap_prot(phys_addr_t addr, unsigned long size,
4cb3cee0 BH	234	unsigned long flags)
4cb3cee0 BH	235	{
1cdab55d BH	236	void *caller = __builtin_return_address(0);
1cdab55d BH	237
a1f242ff	238	/* writeable implies dirty for kernel addresses */
c7d54842	239	if (flags & _PAGE_WRITE)
a1f242ff BH	240	flags \|= _PAGE_DIRTY;
a1f242ff BH	241
ac29c640 AK	242	/* we don't want to let _PAGE_EXEC leak out */
	243	flags &= ~_PAGE_EXEC;
	244	/*
	245	* Force kernel mapping.
	246	*/
ac29c640	247	flags &= ~_PAGE_USER;
812fadcb	248	flags \|= _PAGE_PRIVILEGED;
55052eec	249
4cb3cee0	250	if (ppc_md.ioremap)
1cdab55d BH	251	return ppc_md.ioremap(addr, size, flags, caller);
1cdab55d BH	252	return __ioremap_caller(addr, size, flags, caller);
4cb3cee0 BH	253	}
	254
	255
14cf11af PM	256	/*
	257	* Unmap an IO region and remove it from imalloc'd list.
	258	* Access to IO memory should be serialized by driver.
14cf11af	259	*/
68a64357	260	void __iounmap(volatile void __iomem *token)
14cf11af PM	261	{
	262	void *addr;
	263
f691fa10	264	if (!slab_is_available())
14cf11af PM	265	return;
14cf11af PM	266
3d5134ee BH	267	addr = (void *) ((unsigned long __force)
	268	PCI_FIX_ADDR(token) & PAGE_MASK);
	269	if ((unsigned long)addr < ioremap_bot) {
	270	printk(KERN_WARNING "Attempt to iounmap early bolted mapping"
	271	" at 0x%p\n", addr);
	272	return;
	273	}
	274	vunmap(addr);
14cf11af PM	275	}
14cf11af PM	276
68a64357	277	void iounmap(volatile void __iomem *token)
4cb3cee0 BH	278	{
	279	if (ppc_md.iounmap)
	280	ppc_md.iounmap(token);
	281	else
	282	__iounmap(token);
	283	}
	284
14cf11af	285	EXPORT_SYMBOL(ioremap);
be135f40	286	EXPORT_SYMBOL(ioremap_wc);
40f1ce7f	287	EXPORT_SYMBOL(ioremap_prot);
14cf11af	288	EXPORT_SYMBOL(__ioremap);
a302cb9d	289	EXPORT_SYMBOL(__ioremap_at);
14cf11af	290	EXPORT_SYMBOL(iounmap);
4cb3cee0	291	EXPORT_SYMBOL(__iounmap);
a302cb9d	292	EXPORT_SYMBOL(__iounmap_at);
5c1f6ee9	293
06743521 AK	294	#ifndef __PAGETABLE_PUD_FOLDED
	295	/* 4 level page table */
	296	struct page *pgd_page(pgd_t pgd)
	297	{
	298	if (pgd_huge(pgd))
	299	return pte_page(pgd_pte(pgd));
	300	return virt_to_page(pgd_page_vaddr(pgd));
	301	}
	302	#endif
	303
	304	struct page *pud_page(pud_t pud)
	305	{
	306	if (pud_huge(pud))
	307	return pte_page(pud_pte(pud));
	308	return virt_to_page(pud_page_vaddr(pud));
	309	}
	310
074c2eae AK	311	/*
	312	* For hugepage we have pfn in the pmd, we use PTE_RPN_SHIFT bits for flags
	313	* For PTE page, we have a PTE_FRAG_SIZE (4K) aligned virtual address.
	314	*/
	315	struct page *pmd_page(pmd_t pmd)
	316	{
ebd31197	317	if (pmd_trans_huge(pmd) \|\| pmd_huge(pmd) \|\| pmd_devmap(pmd))
e34aa03c	318	return pte_page(pmd_pte(pmd));
074c2eae AK	319	return virt_to_page(pmd_page_vaddr(pmd));
	320	}
	321
5c1f6ee9 AK	322	#ifdef CONFIG_PPC_64K_PAGES
	323	static pte_t get_from_cache(struct mm_struct mm)
	324	{
	325	void pte_frag, ret;
	326
	327	spin_lock(&mm->page_table_lock);
	328	ret = mm->context.pte_frag;
	329	if (ret) {
	330	pte_frag = ret + PTE_FRAG_SIZE;
	331	/*
	332	* If we have taken up all the fragments mark PTE page NULL
	333	*/
	334	if (((unsigned long)pte_frag & ~PAGE_MASK) == 0)
	335	pte_frag = NULL;
	336	mm->context.pte_frag = pte_frag;
	337	}
	338	spin_unlock(&mm->page_table_lock);
	339	return (pte_t *)ret;
	340	}
	341
	342	static pte_t __alloc_for_cache(struct mm_struct mm, int kernel)
	343	{
	344	void *ret = NULL;
de3b8761 BS	345	struct page *page;
	346
	347	if (!kernel) {
	348	page = alloc_page(PGALLOC_GFP \| __GFP_ACCOUNT);
	349	if (!page)
	350	return NULL;
	351	if (!pgtable_page_ctor(page)) {
	352	__free_page(page);
	353	return NULL;
	354	}
	355	} else {
	356	page = alloc_page(PGALLOC_GFP);
	357	if (!page)
	358	return NULL;
4f804943	359	}
5c1f6ee9 AK	360
	361	ret = page_address(page);
	362	spin_lock(&mm->page_table_lock);
	363	/*
	364	* If we find pgtable_page set, we return
	365	* the allocated page with single fragement
	366	* count.
	367	*/
	368	if (likely(!mm->context.pte_frag)) {
fe896d18	369	set_page_count(page, PTE_FRAG_NR);
5c1f6ee9 AK	370	mm->context.pte_frag = ret + PTE_FRAG_SIZE;
	371	}
	372	spin_unlock(&mm->page_table_lock);
	373
5c1f6ee9 AK	374	return (pte_t *)ret;
	375	}
	376
74701d59	377	pte_t pte_fragment_alloc(struct mm_struct mm, unsigned long vmaddr, int kernel)
5c1f6ee9 AK	378	{
	379	pte_t *pte;
	380
	381	pte = get_from_cache(mm);
	382	if (pte)
	383	return pte;
	384
	385	return __alloc_for_cache(mm, kernel);
	386	}
934828ed	387	#endif /* CONFIG_PPC_64K_PAGES */
5c1f6ee9	388
74701d59	389	void pte_fragment_free(unsigned long *table, int kernel)
5c1f6ee9 AK	390	{
	391	struct page *page = virt_to_page(table);
	392	if (put_page_testzero(page)) {
	393	if (!kernel)
	394	pgtable_page_dtor(page);
2d4894b5	395	free_unref_page(page);
5c1f6ee9 AK	396	}
	397	}
	398
	399	#ifdef CONFIG_SMP
5c1f6ee9 AK	400	void pgtable_free_tlb(struct mmu_gather tlb, void table, int shift)
	401	{
	402	unsigned long pgf = (unsigned long)table;
	403
	404	BUG_ON(shift > MAX_PGTABLE_INDEX_SIZE);
	405	pgf \|= shift;
	406	tlb_remove_table(tlb, (void *)pgf);
	407	}
	408
	409	void __tlb_remove_table(void *_table)
	410	{
	411	void table = (void )((unsigned long)_table & ~MAX_PGTABLE_INDEX_SIZE);
	412	unsigned shift = (unsigned long)_table & MAX_PGTABLE_INDEX_SIZE;
	413
	414	if (!shift)
	415	/* PTE page needs special handling */
74701d59	416	pte_fragment_free(table, 0);
5c1f6ee9 AK	417	else {
	418	BUG_ON(shift > MAX_PGTABLE_INDEX_SIZE);
	419	kmem_cache_free(PGT_CACHE(shift), table);
	420	}
	421	}
	422	#else
	423	void pgtable_free_tlb(struct mmu_gather tlb, void table, int shift)
	424	{
	425	if (!shift) {
	426	/* PTE page needs special handling */
74701d59	427	pte_fragment_free(table, 0);
5c1f6ee9 AK	428	} else {
	429	BUG_ON(shift > MAX_PGTABLE_INDEX_SIZE);
	430	kmem_cache_free(PGT_CACHE(shift), table);
	431	}
	432	}
	433	#endif
9d661958 PM	434
	435	#ifdef CONFIG_PPC_BOOK3S_64
	436	void __init mmu_partition_table_init(void)
	437	{
	438	unsigned long patb_size = 1UL << PATB_SIZE_SHIFT;
1d0761d2	439	unsigned long ptcr;
9d661958 PM	440
	441	BUILD_BUG_ON_MSG((PATB_SIZE_SHIFT > 36), "Partition table size too large.");
	442	partition_tb = __va(memblock_alloc_base(patb_size, patb_size,
	443	MEMBLOCK_ALLOC_ANYWHERE));
	444
	445	/* Initialize the Partition Table with no entries */
	446	memset((void *)partition_tb, 0, patb_size);
	447
	448	/*
	449	* update partition table control register,
	450	* 64 K size.
	451	*/
1d0761d2 AP	452	ptcr = __pa(partition_tb) \| (PATB_SIZE_SHIFT - 12);
	453	mtspr(SPRN_PTCR, ptcr);
	454	powernv_set_nmmu_ptcr(ptcr);
9d661958 PM	455	}
	456
	457	void mmu_partition_table_set_entry(unsigned int lpid, unsigned long dw0,
	458	unsigned long dw1)
	459	{
16ed1416 PM	460	unsigned long old = be64_to_cpu(partition_tb[lpid].patb0);
16ed1416 PM	461
9d661958 PM	462	partition_tb[lpid].patb0 = cpu_to_be64(dw0);
	463	partition_tb[lpid].patb1 = cpu_to_be64(dw1);
	464
16ed1416 PM	465	/*
	466	* Global flush of TLBs and partition table caches for this lpid.
	467	* The type of flush (hash or radix) depends on what the previous
	468	* use of this partition ID was, not the new use.
	469	*/
9d661958	470	asm volatile("ptesync" : : : "memory");
0428491c	471	if (old & PATB_HR) {
16ed1416 PM	472	asm volatile(PPC_TLBIE_5(%0,%1,2,0,1) : :
16ed1416 PM	473	"r" (TLBIEL_INVAL_SET_LPID), "r" (lpid));
62e984dd AK	474	asm volatile(PPC_TLBIE_5(%0,%1,2,1,1) : :
62e984dd AK	475	"r" (TLBIEL_INVAL_SET_LPID), "r" (lpid));
0428491c BS	476	trace_tlbie(lpid, 0, TLBIEL_INVAL_SET_LPID, lpid, 2, 0, 1);
0428491c BS	477	} else {
16ed1416 PM	478	asm volatile(PPC_TLBIE_5(%0,%1,2,0,0) : :
16ed1416 PM	479	"r" (TLBIEL_INVAL_SET_LPID), "r" (lpid));
0428491c BS	480	trace_tlbie(lpid, 0, TLBIEL_INVAL_SET_LPID, lpid, 2, 0, 0);
0428491c BS	481	}
9d661958 PM	482	asm volatile("eieio; tlbsync; ptesync" : : : "memory");
	483	}
	484	EXPORT_SYMBOL_GPL(mmu_partition_table_set_entry);
	485	#endif /* CONFIG_PPC_BOOK3S_64 */
cd65d697 BS	486
	487	#ifdef CONFIG_STRICT_KERNEL_RWX
	488	void mark_rodata_ro(void)
	489	{
	490	if (!mmu_has_feature(MMU_FTR_KERNEL_RO)) {
	491	pr_warn("Warning: Unable to mark rodata read only on this CPU.\n");
	492	return;
	493	}
	494
7614ff32 BS	495	if (radix_enabled())
	496	radix__mark_rodata_ro();
	497	else
cd65d697 BS	498	hash__mark_rodata_ro();
cd65d697 BS	499	}
029d9252 ME	500
	501	void mark_initmem_nx(void)
	502	{
	503	if (radix_enabled())
	504	radix__mark_initmem_nx();
	505	else
	506	hash__mark_initmem_nx();
	507	}
cd65d697	508	#endif