hugetlbpage.c

来自「LINUX 2.6.17.4的源码」· C语言 代码 · 共 1,067 行 · 第 1/2 页

/*
 * PPC64 (POWER4) Huge TLB Page Support for Kernel.
 *
 * Copyright (C) 2003 David Gibson, IBM Corporation.
 *
 * Based on the IA-32 version:
 * Copyright (C) 2002, Rohit Seth <[e-mail]>
 */

#include <linux/init.h>
#include <linux/fs.h>
#include <linux/mm.h>
#include <linux/hugetlb.h>
#include <linux/pagemap.h>
#include <linux/smp_lock.h>
#include <linux/slab.h>
#include <linux/err.h>
#include <linux/sysctl.h>
#include <asm/mman.h>
#include <asm/pgalloc.h>
#include <asm/tlb.h>
#include <asm/tlbflush.h>
#include <asm/mmu_context.h>
#include <asm/machdep.h>
#include <asm/cputable.h>
#include <asm/tlb.h>

#include <linux/sysctl.h>

#define NUM_LOW_AREAS	(0x100000000UL >> SID_SHIFT)
#define NUM_HIGH_AREAS	(PGTABLE_RANGE >> HTLB_AREA_SHIFT)

#ifdef CONFIG_PPC_64K_PAGES
#define HUGEPTE_INDEX_SIZE	(PMD_SHIFT-HPAGE_SHIFT)
#else
#define HUGEPTE_INDEX_SIZE	(PUD_SHIFT-HPAGE_SHIFT)
#endif
#define PTRS_PER_HUGEPTE	(1 << HUGEPTE_INDEX_SIZE)
#define HUGEPTE_TABLE_SIZE	(sizeof(pte_t) << HUGEPTE_INDEX_SIZE)

#define HUGEPD_SHIFT		(HPAGE_SHIFT + HUGEPTE_INDEX_SIZE)
#define HUGEPD_SIZE		(1UL << HUGEPD_SHIFT)
#define HUGEPD_MASK		(~(HUGEPD_SIZE-1))

#define huge_pgtable_cache	(pgtable_cache[HUGEPTE_CACHE_NUM])

/* Flag to mark huge PD pointers.  This means pmd_bad() and pud_bad()
 * will choke on pointers to hugepte tables, which is handy for
 * catching screwups early. */
#define HUGEPD_OK	0x1

typedef struct { unsigned long pd; } hugepd_t;

#define hugepd_none(hpd)	((hpd).pd == 0)

static inline pte_t *hugepd_page(hugepd_t hpd)
{
	BUG_ON(!(hpd.pd & HUGEPD_OK));
	return (pte_t *)(hpd.pd & ~HUGEPD_OK);
}

static inline pte_t *hugepte_offset(hugepd_t *hpdp, unsigned long addr)
{
	unsigned long idx = ((addr >> HPAGE_SHIFT) & (PTRS_PER_HUGEPTE-1));
	pte_t *dir = hugepd_page(*hpdp);

	return dir + idx;
}

static int __hugepte_alloc(struct mm_struct *mm, hugepd_t *hpdp,
			   unsigned long address)
{
	pte_t *new = kmem_cache_alloc(huge_pgtable_cache,
				      GFP_KERNEL|__GFP_REPEAT);

	if (! new)
		return -ENOMEM;

	spin_lock(&mm->page_table_lock);
	if (!hugepd_none(*hpdp))
		kmem_cache_free(huge_pgtable_cache, new);
	else
		hpdp->pd = (unsigned long)new | HUGEPD_OK;
	spin_unlock(&mm->page_table_lock);
	return 0;
}

/* Modelled after find_linux_pte() */
pte_t *huge_pte_offset(struct mm_struct *mm, unsigned long addr)
{
	pgd_t *pg;
	pud_t *pu;

	BUG_ON(! in_hugepage_area(mm->context, addr));

	addr &= HPAGE_MASK;

	pg = pgd_offset(mm, addr);
	if (!pgd_none(*pg)) {
		pu = pud_offset(pg, addr);
		if (!pud_none(*pu)) {
#ifdef CONFIG_PPC_64K_PAGES
			pmd_t *pm;
			pm = pmd_offset(pu, addr);
			if (!pmd_none(*pm))
				return hugepte_offset((hugepd_t *)pm, addr);
#else
			return hugepte_offset((hugepd_t *)pu, addr);
#endif
		}
	}

	return NULL;
}

pte_t *huge_pte_alloc(struct mm_struct *mm, unsigned long addr)
{
	pgd_t *pg;
	pud_t *pu;
	hugepd_t *hpdp = NULL;

	BUG_ON(! in_hugepage_area(mm->context, addr));

	addr &= HPAGE_MASK;

	pg = pgd_offset(mm, addr);
	pu = pud_alloc(mm, pg, addr);

	if (pu) {
#ifdef CONFIG_PPC_64K_PAGES
		pmd_t *pm;
		pm = pmd_alloc(mm, pu, addr);
		if (pm)
			hpdp = (hugepd_t *)pm;
#else
		hpdp = (hugepd_t *)pu;
#endif
	}

	if (! hpdp)
		return NULL;

	if (hugepd_none(*hpdp) && __hugepte_alloc(mm, hpdp, addr))
		return NULL;

	return hugepte_offset(hpdp, addr);
}

static void free_hugepte_range(struct mmu_gather *tlb, hugepd_t *hpdp)
{
	pte_t *hugepte = hugepd_page(*hpdp);

	hpdp->pd = 0;
	tlb->need_flush = 1;
	pgtable_free_tlb(tlb, pgtable_free_cache(hugepte, HUGEPTE_CACHE_NUM,
						 HUGEPTE_TABLE_SIZE-1));
}

#ifdef CONFIG_PPC_64K_PAGES
static void hugetlb_free_pmd_range(struct mmu_gather *tlb, pud_t *pud,
				   unsigned long addr, unsigned long end,
				   unsigned long floor, unsigned long ceiling)
{
	pmd_t *pmd;
	unsigned long next;
	unsigned long start;

	start = addr;
	pmd = pmd_offset(pud, addr);
	do {
		next = pmd_addr_end(addr, end);
		if (pmd_none(*pmd))
			continue;
		free_hugepte_range(tlb, (hugepd_t *)pmd);
	} while (pmd++, addr = next, addr != end);

	start &= PUD_MASK;
	if (start < floor)
		return;
	if (ceiling) {
		ceiling &= PUD_MASK;
		if (!ceiling)
			return;
	}
	if (end - 1 > ceiling - 1)
		return;

	pmd = pmd_offset(pud, start);
	pud_clear(pud);
	pmd_free_tlb(tlb, pmd);
}
#endif

static void hugetlb_free_pud_range(struct mmu_gather *tlb, pgd_t *pgd,
				   unsigned long addr, unsigned long end,
				   unsigned long floor, unsigned long ceiling)
{
	pud_t *pud;
	unsigned long next;
	unsigned long start;

	start = addr;
	pud = pud_offset(pgd, addr);
	do {
		next = pud_addr_end(addr, end);
#ifdef CONFIG_PPC_64K_PAGES
		if (pud_none_or_clear_bad(pud))
			continue;
		hugetlb_free_pmd_range(tlb, pud, addr, next, floor, ceiling);
#else
		if (pud_none(*pud))
			continue;
		free_hugepte_range(tlb, (hugepd_t *)pud);
#endif
	} while (pud++, addr = next, addr != end);

	start &= PGDIR_MASK;
	if (start < floor)
		return;
	if (ceiling) {
		ceiling &= PGDIR_MASK;
		if (!ceiling)
			return;
	}
	if (end - 1 > ceiling - 1)
		return;

	pud = pud_offset(pgd, start);
	pgd_clear(pgd);
	pud_free_tlb(tlb, pud);
}

/*
 * This function frees user-level page tables of a process.
 *
 * Must be called with pagetable lock held.
 */
void hugetlb_free_pgd_range(struct mmu_gather **tlb,
			    unsigned long addr, unsigned long end,
			    unsigned long floor, unsigned long ceiling)
{
	pgd_t *pgd;
	unsigned long next;
	unsigned long start;

	/*
	 * Comments below take from the normal free_pgd_range().  They
	 * apply here too.  The tests against HUGEPD_MASK below are
	 * essential, because we *don't* test for this at the bottom
	 * level.  Without them we'll attempt to free a hugepte table
	 * when we unmap just part of it, even if there are other
	 * active mappings using it.
	 *
	 * The next few lines have given us lots of grief...
	 *
	 * Why are we testing HUGEPD* at this top level?  Because
	 * often there will be no work to do at all, and we'd prefer
	 * not to go all the way down to the bottom just to discover
	 * that.
	 *
	 * Why all these "- 1"s?  Because 0 represents both the bottom
	 * of the address space and the top of it (using -1 for the
	 * top wouldn't help much: the masks would do the wrong thing).
	 * The rule is that addr 0 and floor 0 refer to the bottom of
	 * the address space, but end 0 and ceiling 0 refer to the top
	 * Comparisons need to use "end - 1" and "ceiling - 1" (though
	 * that end 0 case should be mythical).
	 *
	 * Wherever addr is brought up or ceiling brought down, we
	 * must be careful to reject "the opposite 0" before it
	 * confuses the subsequent tests.  But what about where end is
	 * brought down by HUGEPD_SIZE below? no, end can't go down to
	 * 0 there.
	 *
	 * Whereas we round start (addr) and ceiling down, by different
	 * masks at different levels, in order to test whether a table
	 * now has no other vmas using it, so can be freed, we don't
	 * bother to round floor or end up - the tests don't need that.
	 */

	addr &= HUGEPD_MASK;
	if (addr < floor) {
		addr += HUGEPD_SIZE;
		if (!addr)
			return;
	}
	if (ceiling) {
		ceiling &= HUGEPD_MASK;
		if (!ceiling)
			return;
	}
	if (end - 1 > ceiling - 1)
		end -= HUGEPD_SIZE;
	if (addr > end - 1)
		return;

	start = addr;
	pgd = pgd_offset((*tlb)->mm, addr);
	do {
		BUG_ON(! in_hugepage_area((*tlb)->mm->context, addr));
		next = pgd_addr_end(addr, end);
		if (pgd_none_or_clear_bad(pgd))
			continue;
		hugetlb_free_pud_range(*tlb, pgd, addr, next, floor, ceiling);
	} while (pgd++, addr = next, addr != end);
}

void set_huge_pte_at(struct mm_struct *mm, unsigned long addr,
		     pte_t *ptep, pte_t pte)
{
	if (pte_present(*ptep)) {
		/* We open-code pte_clear because we need to pass the right
		 * argument to hpte_update (huge / !huge)
		 */
		unsigned long old = pte_update(ptep, ~0UL);
		if (old & _PAGE_HASHPTE)
			hpte_update(mm, addr & HPAGE_MASK, ptep, old, 1);
		flush_tlb_pending();
	}
	*ptep = __pte(pte_val(pte) & ~_PAGE_HPTEFLAGS);
}

pte_t huge_ptep_get_and_clear(struct mm_struct *mm, unsigned long addr,
			      pte_t *ptep)
{
	unsigned long old = pte_update(ptep, ~0UL);

	if (old & _PAGE_HASHPTE)
		hpte_update(mm, addr & HPAGE_MASK, ptep, old, 1);
	*ptep = __pte(0);

	return __pte(old);
}

struct slb_flush_info {
	struct mm_struct *mm;
	u16 newareas;
};

static void flush_low_segments(void *parm)
{
	struct slb_flush_info *fi = parm;
	unsigned long i;

	BUILD_BUG_ON((sizeof(fi->newareas)*8) != NUM_LOW_AREAS);

	if (current->active_mm != fi->mm)
		return;

	/* Only need to do anything if this CPU is working in the same
	 * mm as the one which has changed */

	/* update the paca copy of the context struct */
	get_paca()->context = current->active_mm->context;

	asm volatile("isync" : : : "memory");
	for (i = 0; i < NUM_LOW_AREAS; i++) {
		if (! (fi->newareas & (1U << i)))
			continue;
		asm volatile("slbie %0"
			     : : "r" ((i << SID_SHIFT) | SLBIE_C));
	}
	asm volatile("isync" : : : "memory");
}

static void flush_high_segments(void *parm)
{
	struct slb_flush_info *fi = parm;
	unsigned long i, j;


	BUILD_BUG_ON((sizeof(fi->newareas)*8) != NUM_HIGH_AREAS);

	if (current->active_mm != fi->mm)
		return;

	/* Only need to do anything if this CPU is working in the same
	 * mm as the one which has changed */

	/* update the paca copy of the context struct */
	get_paca()->context = current->active_mm->context;

	asm volatile("isync" : : : "memory");
	for (i = 0; i < NUM_HIGH_AREAS; i++) {
		if (! (fi->newareas & (1U << i)))
			continue;
		for (j = 0; j < (1UL << (HTLB_AREA_SHIFT-SID_SHIFT)); j++)
			asm volatile("slbie %0"
				     :: "r" (((i << HTLB_AREA_SHIFT)
					      + (j << SID_SHIFT)) | SLBIE_C));
	}
	asm volatile("isync" : : : "memory");
}

static int prepare_low_area_for_htlb(struct mm_struct *mm, unsigned long area)
{
	unsigned long start = area << SID_SHIFT;
	unsigned long end = (area+1) << SID_SHIFT;
	struct vm_area_struct *vma;

	BUG_ON(area >= NUM_LOW_AREAS);

	/* Check no VMAs are in the region */
	vma = find_vma(mm, start);
	if (vma && (vma->vm_start < end))
		return -EBUSY;

	return 0;
}

static int prepare_high_area_for_htlb(struct mm_struct *mm, unsigned long area)
{
	unsigned long start = area << HTLB_AREA_SHIFT;
	unsigned long end = (area+1) << HTLB_AREA_SHIFT;
	struct vm_area_struct *vma;

	BUG_ON(area >= NUM_HIGH_AREAS);

	/* Hack, so that each addresses is controlled by exactly one
	 * of the high or low area bitmaps, the first high area starts
	 * at 4GB, not 0 */
	if (start == 0)
		start = 0x100000000UL;

	/* Check no VMAs are in the region */
	vma = find_vma(mm, start);
	if (vma && (vma->vm_start < end))
		return -EBUSY;

	return 0;
}

static int open_low_hpage_areas(struct mm_struct *mm, u16 newareas)
{
	unsigned long i;
	struct slb_flush_info fi;

	BUILD_BUG_ON((sizeof(newareas)*8) != NUM_LOW_AREAS);
	BUILD_BUG_ON((sizeof(mm->context.low_htlb_areas)*8) != NUM_LOW_AREAS);

	newareas &= ~(mm->context.low_htlb_areas);
	if (! newareas)
		return 0; /* The segments we want are already open */

	for (i = 0; i < NUM_LOW_AREAS; i++)
		if ((1 << i) & newareas)
			if (prepare_low_area_for_htlb(mm, i) != 0)
				return -EBUSY;

	mm->context.low_htlb_areas |= newareas;

	/* the context change must make it to memory before the flush,
	 * so that further SLB misses do the right thing. */
	mb();

	fi.mm = mm;
	fi.newareas = newareas;
	on_each_cpu(flush_low_segments, &fi, 0, 1);

	return 0;
}

static int open_high_hpage_areas(struct mm_struct *mm, u16 newareas)
{
	struct slb_flush_info fi;
	unsigned long i;

	BUILD_BUG_ON((sizeof(newareas)*8) != NUM_HIGH_AREAS);
	BUILD_BUG_ON((sizeof(mm->context.high_htlb_areas)*8)
		     != NUM_HIGH_AREAS);

	newareas &= ~(mm->context.high_htlb_areas);
	if (! newareas)
		return 0; /* The areas we want are already open */

	for (i = 0; i < NUM_HIGH_AREAS; i++)
		if ((1 << i) & newareas)
			if (prepare_high_area_for_htlb(mm, i) != 0)
				return -EBUSY;

	mm->context.high_htlb_areas |= newareas;

	/* update the paca copy of the context struct */
	get_paca()->context = mm->context;

	/* the context change must make it to memory before the flush,
	 * so that further SLB misses do the right thing. */
	mb();

	fi.mm = mm;
	fi.newareas = newareas;
	on_each_cpu(flush_high_segments, &fi, 0, 1);

	return 0;
}

int prepare_hugepage_range(unsigned long addr, unsigned long len)
{
	int err = 0;

	if ( (addr+len) < addr )
		return -EINVAL;

	if (addr < 0x100000000UL)
		err = open_low_hpage_areas(current->mm,
					  LOW_ESID_MASK(addr, len));
	if ((addr + len) > 0x100000000UL)
		err = open_high_hpage_areas(current->mm,
					    HTLB_AREA_MASK(addr, len));
	if (err) {
		printk(KERN_DEBUG "prepare_hugepage_range(%lx, %lx)"
		       " failed (lowmask: 0x%04hx, highmask: 0x%04hx)\n",
		       addr, len,
		       LOW_ESID_MASK(addr, len), HTLB_AREA_MASK(addr, len));
		return err;
	}

	return 0;
}

struct page *
follow_huge_addr(struct mm_struct *mm, unsigned long address, int write)
{
	pte_t *ptep;
	struct page *page;

	if (! in_hugepage_area(mm->context, address))
		return ERR_PTR(-EINVAL);

	ptep = huge_pte_offset(mm, address);
	page = pte_page(*ptep);
	if (page)
		page += (address % HPAGE_SIZE) / PAGE_SIZE;