[linux-2.6-block.git] / arch / s390 / mm / hugetlbpage.c

/*
 *  IBM System z Huge TLB Page Support for Kernel.
 *
 *    Copyright IBM Corp. 2007
 *    Author(s): Gerald Schaefer <gerald.schaefer@de.ibm.com>
 */

#include <linux/mm.h>
#include <linux/hugetlb.h>

static inline pmd_t __pte_to_pmd(pte_t pte)
{
	pmd_t pmd;

	/*
	 * Convert encoding		  pte bits	   pmd bits
	 *				lIR.uswrdy.p	dy..R...I...wr
	 * empty			010.000000.0 -> 00..0...1...00
	 * prot-none, clean, old	111.000000.1 -> 00..1...1...00
	 * prot-none, clean, young	111.000001.1 -> 01..1...1...00
	 * prot-none, dirty, old	111.000010.1 -> 10..1...1...00
	 * prot-none, dirty, young	111.000011.1 -> 11..1...1...00
	 * read-only, clean, old	111.000100.1 -> 00..1...1...01
	 * read-only, clean, young	101.000101.1 -> 01..1...0...01
	 * read-only, dirty, old	111.000110.1 -> 10..1...1...01
	 * read-only, dirty, young	101.000111.1 -> 11..1...0...01
	 * read-write, clean, old	111.001100.1 -> 00..1...1...11
	 * read-write, clean, young	101.001101.1 -> 01..1...0...11
	 * read-write, dirty, old	110.001110.1 -> 10..0...1...11
	 * read-write, dirty, young	100.001111.1 -> 11..0...0...11
	 * HW-bits: R read-only, I invalid
	 * SW-bits: p present, y young, d dirty, r read, w write, s special,
	 *	    u unused, l large
	 */
	if (pte_present(pte)) {
		pmd_val(pmd) = pte_val(pte) & PAGE_MASK;
		pmd_val(pmd) |= (pte_val(pte) & _PAGE_READ) >> 4;
		pmd_val(pmd) |= (pte_val(pte) & _PAGE_WRITE) >> 4;
		pmd_val(pmd) |=	(pte_val(pte) & _PAGE_INVALID) >> 5;
		pmd_val(pmd) |= (pte_val(pte) & _PAGE_PROTECT);
		pmd_val(pmd) |= (pte_val(pte) & _PAGE_DIRTY) << 10;
		pmd_val(pmd) |= (pte_val(pte) & _PAGE_YOUNG) << 10;
	} else
		pmd_val(pmd) = _SEGMENT_ENTRY_INVALID;
	return pmd;
}

static inline pte_t __pmd_to_pte(pmd_t pmd)
{
	pte_t pte;

	/*
	 * Convert encoding		   pmd bits	    pte bits
	 *				dy..R...I...wr	  lIR.uswrdy.p
	 * empty			00..0...1...00 -> 010.000000.0
	 * prot-none, clean, old	00..1...1...00 -> 111.000000.1
	 * prot-none, clean, young	01..1...1...00 -> 111.000001.1
	 * prot-none, dirty, old	10..1...1...00 -> 111.000010.1
	 * prot-none, dirty, young	11..1...1...00 -> 111.000011.1
	 * read-only, clean, old	00..1...1...01 -> 111.000100.1
	 * read-only, clean, young	01..1...0...01 -> 101.000101.1
	 * read-only, dirty, old	10..1...1...01 -> 111.000110.1
	 * read-only, dirty, young	11..1...0...01 -> 101.000111.1
	 * read-write, clean, old	00..1...1...11 -> 111.001100.1
	 * read-write, clean, young	01..1...0...11 -> 101.001101.1
	 * read-write, dirty, old	10..0...1...11 -> 110.001110.1
	 * read-write, dirty, young	11..0...0...11 -> 100.001111.1
	 * HW-bits: R read-only, I invalid
	 * SW-bits: p present, y young, d dirty, r read, w write, s special,
	 *	    u unused, l large
	 */
	if (pmd_present(pmd)) {
		pte_val(pte) = pmd_val(pmd) & _SEGMENT_ENTRY_ORIGIN_LARGE;
		pte_val(pte) |= _PAGE_LARGE | _PAGE_PRESENT;
		pte_val(pte) |= (pmd_val(pmd) & _SEGMENT_ENTRY_READ) << 4;
		pte_val(pte) |= (pmd_val(pmd) & _SEGMENT_ENTRY_WRITE) << 4;
		pte_val(pte) |= (pmd_val(pmd) & _SEGMENT_ENTRY_INVALID) << 5;
		pte_val(pte) |= (pmd_val(pmd) & _SEGMENT_ENTRY_PROTECT);
		pte_val(pte) |= (pmd_val(pmd) & _SEGMENT_ENTRY_DIRTY) >> 10;
		pte_val(pte) |= (pmd_val(pmd) & _SEGMENT_ENTRY_YOUNG) >> 10;
	} else
		pte_val(pte) = _PAGE_INVALID;
	return pte;
}

void set_huge_pte_at(struct mm_struct *mm, unsigned long addr,
		     pte_t *ptep, pte_t pte)
{
	pmd_t pmd;

	pmd = __pte_to_pmd(pte);
	if (!MACHINE_HAS_HPAGE) {
		/* Emulated huge ptes loose the dirty and young bit */
		pmd_val(pmd) &= ~_SEGMENT_ENTRY_ORIGIN;
		pmd_val(pmd) |= pte_page(pte)[1].index;
	} else
		pmd_val(pmd) |= _SEGMENT_ENTRY_LARGE;
	*(pmd_t *) ptep = pmd;
}

pte_t huge_ptep_get(pte_t *ptep)
{
	unsigned long origin;
	pmd_t pmd;

	pmd = *(pmd_t *) ptep;
	if (!MACHINE_HAS_HPAGE && pmd_present(pmd)) {
		origin = pmd_val(pmd) & _SEGMENT_ENTRY_ORIGIN;
		pmd_val(pmd) &= ~_SEGMENT_ENTRY_ORIGIN;
		pmd_val(pmd) |= *(unsigned long *) origin;
		/* Emulated huge ptes are young and dirty by definition */
		pmd_val(pmd) |= _SEGMENT_ENTRY_YOUNG | _SEGMENT_ENTRY_DIRTY;
	}
	return __pmd_to_pte(pmd);
}

pte_t huge_ptep_get_and_clear(struct mm_struct *mm,
			      unsigned long addr, pte_t *ptep)
{
	pmd_t *pmdp = (pmd_t *) ptep;
	pte_t pte = huge_ptep_get(ptep);

	pmdp_flush_direct(mm, addr, pmdp);
	pmd_val(*pmdp) = _SEGMENT_ENTRY_EMPTY;
	return pte;
}

int arch_prepare_hugepage(struct page *page)
{
	unsigned long addr = page_to_phys(page);
	pte_t pte;
	pte_t *ptep;
	int i;

	if (MACHINE_HAS_HPAGE)
		return 0;

	ptep = (pte_t *) pte_alloc_one(&init_mm, addr);
	if (!ptep)
		return -ENOMEM;

	pte_val(pte) = addr;
	for (i = 0; i < PTRS_PER_PTE; i++) {
		set_pte_at(&init_mm, addr + i * PAGE_SIZE, ptep + i, pte);
		pte_val(pte) += PAGE_SIZE;
	}
	page[1].index = (unsigned long) ptep;
	return 0;
}

void arch_release_hugepage(struct page *page)
{
	pte_t *ptep;

	if (MACHINE_HAS_HPAGE)
		return;

	ptep = (pte_t *) page[1].index;
	if (!ptep)
		return;
	clear_table((unsigned long *) ptep, _PAGE_INVALID,
		    PTRS_PER_PTE * sizeof(pte_t));
	page_table_free(&init_mm, (unsigned long *) ptep);
	page[1].index = 0;
}

pte_t *huge_pte_alloc(struct mm_struct *mm,
			unsigned long addr, unsigned long sz)
{
	pgd_t *pgdp;
	pud_t *pudp;
	pmd_t *pmdp = NULL;

	pgdp = pgd_offset(mm, addr);
	pudp = pud_alloc(mm, pgdp, addr);
	if (pudp)
		pmdp = pmd_alloc(mm, pudp, addr);
	return (pte_t *) pmdp;
}

pte_t *huge_pte_offset(struct mm_struct *mm, unsigned long addr)
{
	pgd_t *pgdp;
	pud_t *pudp;
	pmd_t *pmdp = NULL;

	pgdp = pgd_offset(mm, addr);
	if (pgd_present(*pgdp)) {
		pudp = pud_offset(pgdp, addr);
		if (pud_present(*pudp))
			pmdp = pmd_offset(pudp, addr);
	}
	return (pte_t *) pmdp;
}

int huge_pmd_unshare(struct mm_struct *mm, unsigned long *addr, pte_t *ptep)
{
	return 0;
}

int pmd_huge(pmd_t pmd)
{
	if (!MACHINE_HAS_HPAGE)
		return 0;

	return !!(pmd_val(pmd) & _SEGMENT_ENTRY_LARGE);
}

int pud_huge(pud_t pud)
{
	return 0;
}
Commit	Line	Data
53492b1d GS	1	/*
	2	* IBM System z Huge TLB Page Support for Kernel.
	3	*
a53c8fab	4	* Copyright IBM Corp. 2007
53492b1d GS	5	* Author(s): Gerald Schaefer <gerald.schaefer@de.ibm.com>
	6	*/
	7
	8	#include <linux/mm.h>
	9	#include <linux/hugetlb.h>
	10
e5098611 MS	11	static inline pmd_t __pte_to_pmd(pte_t pte)
e5098611 MS	12	{
e5098611 MS	13	pmd_t pmd;
	14
	15	/*
152125b7	16	* Convert encoding pte bits pmd bits
a1c843b8 MS	17	* lIR.uswrdy.p dy..R...I...wr
	18	* empty 010.000000.0 -> 00..0...1...00
	19	* prot-none, clean, old 111.000000.1 -> 00..1...1...00
	20	* prot-none, clean, young 111.000001.1 -> 01..1...1...00
	21	* prot-none, dirty, old 111.000010.1 -> 10..1...1...00
	22	* prot-none, dirty, young 111.000011.1 -> 11..1...1...00
	23	* read-only, clean, old 111.000100.1 -> 00..1...1...01
	24	* read-only, clean, young 101.000101.1 -> 01..1...0...01
	25	* read-only, dirty, old 111.000110.1 -> 10..1...1...01
	26	* read-only, dirty, young 101.000111.1 -> 11..1...0...01
	27	* read-write, clean, old 111.001100.1 -> 00..1...1...11
	28	* read-write, clean, young 101.001101.1 -> 01..1...0...11
	29	* read-write, dirty, old 110.001110.1 -> 10..0...1...11
	30	* read-write, dirty, young 100.001111.1 -> 11..0...0...11
	31	* HW-bits: R read-only, I invalid
	32	* SW-bits: p present, y young, d dirty, r read, w write, s special,
	33	* u unused, l large
e5098611 MS	34	*/
	35	if (pte_present(pte)) {
	36	pmd_val(pmd) = pte_val(pte) & PAGE_MASK;
152125b7 MS	37	pmd_val(pmd) \|= (pte_val(pte) & _PAGE_READ) >> 4;
	38	pmd_val(pmd) \|= (pte_val(pte) & _PAGE_WRITE) >> 4;
	39	pmd_val(pmd) \|= (pte_val(pte) & _PAGE_INVALID) >> 5;
	40	pmd_val(pmd) \|= (pte_val(pte) & _PAGE_PROTECT);
	41	pmd_val(pmd) \|= (pte_val(pte) & _PAGE_DIRTY) << 10;
	42	pmd_val(pmd) \|= (pte_val(pte) & _PAGE_YOUNG) << 10;
e5098611 MS	43	} else
	44	pmd_val(pmd) = _SEGMENT_ENTRY_INVALID;
	45	return pmd;
	46	}
	47
	48	static inline pte_t __pmd_to_pte(pmd_t pmd)
	49	{
	50	pte_t pte;
	51
	52	/*
152125b7	53	* Convert encoding pmd bits pte bits
a1c843b8 MS	54	* dy..R...I...wr lIR.uswrdy.p
	55	* empty 00..0...1...00 -> 010.000000.0
	56	* prot-none, clean, old 00..1...1...00 -> 111.000000.1
	57	* prot-none, clean, young 01..1...1...00 -> 111.000001.1
	58	* prot-none, dirty, old 10..1...1...00 -> 111.000010.1
	59	* prot-none, dirty, young 11..1...1...00 -> 111.000011.1
	60	* read-only, clean, old 00..1...1...01 -> 111.000100.1
	61	* read-only, clean, young 01..1...0...01 -> 101.000101.1
	62	* read-only, dirty, old 10..1...1...01 -> 111.000110.1
	63	* read-only, dirty, young 11..1...0...01 -> 101.000111.1
	64	* read-write, clean, old 00..1...1...11 -> 111.001100.1
	65	* read-write, clean, young 01..1...0...11 -> 101.001101.1
	66	* read-write, dirty, old 10..0...1...11 -> 110.001110.1
	67	* read-write, dirty, young 11..0...0...11 -> 100.001111.1
	68	* HW-bits: R read-only, I invalid
	69	* SW-bits: p present, y young, d dirty, r read, w write, s special,
	70	* u unused, l large
e5098611 MS	71	*/
e5098611 MS	72	if (pmd_present(pmd)) {
152125b7 MS	73	pte_val(pte) = pmd_val(pmd) & _SEGMENT_ENTRY_ORIGIN_LARGE;
	74	pte_val(pte) \|= _PAGE_LARGE \| _PAGE_PRESENT;
	75	pte_val(pte) \|= (pmd_val(pmd) & _SEGMENT_ENTRY_READ) << 4;
	76	pte_val(pte) \|= (pmd_val(pmd) & _SEGMENT_ENTRY_WRITE) << 4;
	77	pte_val(pte) \|= (pmd_val(pmd) & _SEGMENT_ENTRY_INVALID) << 5;
	78	pte_val(pte) \|= (pmd_val(pmd) & _SEGMENT_ENTRY_PROTECT);
b7d14f3a MS	79	pte_val(pte) \|= (pmd_val(pmd) & _SEGMENT_ENTRY_DIRTY) >> 10;
b7d14f3a MS	80	pte_val(pte) \|= (pmd_val(pmd) & _SEGMENT_ENTRY_YOUNG) >> 10;
e5098611 MS	81	} else
	82	pte_val(pte) = _PAGE_INVALID;
	83	return pte;
	84	}
53492b1d GS	85
53492b1d GS	86	void set_huge_pte_at(struct mm_struct *mm, unsigned long addr,
e5098611	87	pte_t *ptep, pte_t pte)
53492b1d	88	{
e5098611	89	pmd_t pmd;
53492b1d	90
e5098611	91	pmd = __pte_to_pmd(pte);
53492b1d	92	if (!MACHINE_HAS_HPAGE) {
152125b7	93	/* Emulated huge ptes loose the dirty and young bit */
e5098611 MS	94	pmd_val(pmd) &= ~_SEGMENT_ENTRY_ORIGIN;
	95	pmd_val(pmd) \|= pte_page(pte)[1].index;
	96	} else
6a5c1482	97	pmd_val(pmd) \|= _SEGMENT_ENTRY_LARGE;
e5098611 MS	98	(pmd_t ) ptep = pmd;
	99	}
	100
	101	pte_t huge_ptep_get(pte_t *ptep)
	102	{
	103	unsigned long origin;
	104	pmd_t pmd;
	105
	106	pmd = (pmd_t ) ptep;
	107	if (!MACHINE_HAS_HPAGE && pmd_present(pmd)) {
	108	origin = pmd_val(pmd) & _SEGMENT_ENTRY_ORIGIN;
	109	pmd_val(pmd) &= ~_SEGMENT_ENTRY_ORIGIN;
	110	pmd_val(pmd) \|= (unsigned long ) origin;
152125b7 MS	111	/* Emulated huge ptes are young and dirty by definition */
152125b7 MS	112	pmd_val(pmd) \|= _SEGMENT_ENTRY_YOUNG \| _SEGMENT_ENTRY_DIRTY;
53492b1d	113	}
e5098611 MS	114	return __pmd_to_pte(pmd);
	115	}
	116
	117	pte_t huge_ptep_get_and_clear(struct mm_struct *mm,
	118	unsigned long addr, pte_t *ptep)
	119	{
	120	pmd_t pmdp = (pmd_t ) ptep;
	121	pte_t pte = huge_ptep_get(ptep);
53492b1d	122
1b948d6c	123	pmdp_flush_direct(mm, addr, pmdp);
e5098611 MS	124	pmd_val(*pmdp) = _SEGMENT_ENTRY_EMPTY;
e5098611 MS	125	return pte;
53492b1d GS	126	}
	127
	128	int arch_prepare_hugepage(struct page *page)
	129	{
	130	unsigned long addr = page_to_phys(page);
	131	pte_t pte;
	132	pte_t *ptep;
	133	int i;
	134
	135	if (MACHINE_HAS_HPAGE)
	136	return 0;
	137
e5992f2e	138	ptep = (pte_t *) pte_alloc_one(&init_mm, addr);
53492b1d GS	139	if (!ptep)
	140	return -ENOMEM;
	141
106c992a	142	pte_val(pte) = addr;
53492b1d GS	143	for (i = 0; i < PTRS_PER_PTE; i++) {
	144	set_pte_at(&init_mm, addr + i * PAGE_SIZE, ptep + i, pte);
	145	pte_val(pte) += PAGE_SIZE;
	146	}
	147	page[1].index = (unsigned long) ptep;
	148	return 0;
	149	}
	150
	151	void arch_release_hugepage(struct page *page)
	152	{
	153	pte_t *ptep;
	154
	155	if (MACHINE_HAS_HPAGE)
	156	return;
	157
	158	ptep = (pte_t *) page[1].index;
	159	if (!ptep)
	160	return;
e5098611	161	clear_table((unsigned long *) ptep, _PAGE_INVALID,
a686425b	162	PTRS_PER_PTE * sizeof(pte_t));
80217147	163	page_table_free(&init_mm, (unsigned long *) ptep);
53492b1d GS	164	page[1].index = 0;
	165	}
	166
a5516438 AK	167	pte_t huge_pte_alloc(struct mm_struct mm,
a5516438 AK	168	unsigned long addr, unsigned long sz)
53492b1d GS	169	{
	170	pgd_t *pgdp;
	171	pud_t *pudp;
	172	pmd_t *pmdp = NULL;
	173
	174	pgdp = pgd_offset(mm, addr);
	175	pudp = pud_alloc(mm, pgdp, addr);
	176	if (pudp)
	177	pmdp = pmd_alloc(mm, pudp, addr);
	178	return (pte_t *) pmdp;
	179	}
	180
	181	pte_t huge_pte_offset(struct mm_struct mm, unsigned long addr)
	182	{
	183	pgd_t *pgdp;
	184	pud_t *pudp;
	185	pmd_t *pmdp = NULL;
	186
	187	pgdp = pgd_offset(mm, addr);
	188	if (pgd_present(*pgdp)) {
	189	pudp = pud_offset(pgdp, addr);
	190	if (pud_present(*pudp))
	191	pmdp = pmd_offset(pudp, addr);
	192	}
	193	return (pte_t *) pmdp;
	194	}
	195
	196	int huge_pmd_unshare(struct mm_struct mm, unsigned long addr, pte_t *ptep)
	197	{
	198	return 0;
	199	}
	200
53492b1d GS	201	int pmd_huge(pmd_t pmd)
	202	{
	203	if (!MACHINE_HAS_HPAGE)
	204	return 0;
	205
	206	return !!(pmd_val(pmd) & _SEGMENT_ENTRY_LARGE);
	207	}
	208
ceb86879 AK	209	int pud_huge(pud_t pud)
	210	{
	211	return 0;
	212	}