summaryrefslogtreecommitdiff
path: root/arch/ppc64/mm/hugetlbpage.c
blob: b4ab766f59800bcc66774d6ff70c2accc390f835 (plain)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
/*
 * 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 <rohit.seth@intel.com>
 */

#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	HUGEPGDIR_SHIFT		(HPAGE_SHIFT + PAGE_SHIFT - 3)
#define HUGEPGDIR_SIZE		(1UL << HUGEPGDIR_SHIFT)
#define HUGEPGDIR_MASK		(~(HUGEPGDIR_SIZE-1))

#define HUGEPTE_INDEX_SIZE	9
#define HUGEPGD_INDEX_SIZE	10

#define PTRS_PER_HUGEPTE	(1 << HUGEPTE_INDEX_SIZE)
#define PTRS_PER_HUGEPGD	(1 << HUGEPGD_INDEX_SIZE)

static inline int hugepgd_index(unsigned long addr)
{
	return (addr & ~REGION_MASK) >> HUGEPGDIR_SHIFT;
}

static pud_t *hugepgd_offset(struct mm_struct *mm, unsigned long addr)
{
	int index;

	if (! mm->context.huge_pgdir)
		return NULL;


	index = hugepgd_index(addr);
	BUG_ON(index >= PTRS_PER_HUGEPGD);
	return (pud_t *)(mm->context.huge_pgdir + index);
}

static inline pte_t *hugepte_offset(pud_t *dir, unsigned long addr)
{
	int index;

	if (pud_none(*dir))
		return NULL;

	index = (addr >> HPAGE_SHIFT) % PTRS_PER_HUGEPTE;
	return (pte_t *)pud_page(*dir) + index;
}

static pud_t *hugepgd_alloc(struct mm_struct *mm, unsigned long addr)
{
	BUG_ON(! in_hugepage_area(mm->context, addr));

	if (! mm->context.huge_pgdir) {
		pgd_t *new;
		spin_unlock(&mm->page_table_lock);
		/* Don't use pgd_alloc(), because we want __GFP_REPEAT */
		new = kmem_cache_alloc(zero_cache, GFP_KERNEL | __GFP_REPEAT);
		BUG_ON(memcmp(new, empty_zero_page, PAGE_SIZE));
		spin_lock(&mm->page_table_lock);

		/*
		 * Because we dropped the lock, we should re-check the
		 * entry, as somebody else could have populated it..
		 */
		if (mm->context.huge_pgdir)
			pgd_free(new);
		else
			mm->context.huge_pgdir = new;
	}
	return hugepgd_offset(mm, addr);
}

static pte_t *hugepte_alloc(struct mm_struct *mm, pud_t *dir, unsigned long addr)
{
	if (! pud_present(*dir)) {
		pte_t *new;

		spin_unlock(&mm->page_table_lock);
		new = kmem_cache_alloc(zero_cache, GFP_KERNEL | __GFP_REPEAT);
		BUG_ON(memcmp(new, empty_zero_page, PAGE_SIZE));
		spin_lock(&mm->page_table_lock);
		/*
		 * Because we dropped the lock, we should re-check the
		 * entry, as somebody else could have populated it..
		 */
		if (pud_present(*dir)) {
			if (new)
				kmem_cache_free(zero_cache, new);
		} else {
			struct page *ptepage;

			if (! new)
				return NULL;
			ptepage = virt_to_page(new);
			ptepage->mapping = (void *) mm;
			ptepage->index = addr & HUGEPGDIR_MASK;
			pud_populate(mm, dir, new);
		}
	}

	return hugepte_offset(dir, addr);
}

pte_t *huge_pte_offset(struct mm_struct *mm, unsigned long addr)
{
	pud_t *pud;

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

	pud = hugepgd_offset(mm, addr);
	if (! pud)
		return NULL;

	return hugepte_offset(pud, addr);
}

pte_t *huge_pte_alloc(struct mm_struct *mm, unsigned long addr)
{
	pud_t *pud;

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

	pud = hugepgd_alloc(mm, addr);
	if (! pud)
		return NULL;

	return hugepte_alloc(mm, pud, addr);
}

/*
 * This function checks for proper alignment of input addr and len parameters.
 */
int is_aligned_hugepage_range(unsigned long addr, unsigned long len)
{
	if (len & ~HPAGE_MASK)
		return -EINVAL;
	if (addr & ~HPAGE_MASK)
		return -EINVAL;
	if (! (within_hugepage_low_range(addr, len)
	       || within_hugepage_high_range(addr, len)) )
		return -EINVAL;
	return 0;
}

static void flush_segments(void *parm)
{
	u16 segs = (unsigned long) parm;
	unsigned long i;

	asm volatile("isync" : : : "memory");

	for (i = 0; i < 16; i++) {
		if (! (segs & (1U << i)))
			continue;
		asm volatile("slbie %0" : : "r" (i << SID_SHIFT));
	}

	asm volatile("isync" : : : "memory");
}

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

	BUG_ON(seg >= 16);

	/* 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_segs(struct mm_struct *mm, u16 newsegs)
{
	unsigned long i;

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

	for (i = 0; i < 16; i++)
		if ((1 << i) & newsegs)
			if (prepare_low_seg_for_htlb(mm, i) != 0)
				return -EBUSY;

	mm->context.htlb_segs |= newsegs;

	/* 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();
	on_each_cpu(flush_segments, (void *)(unsigned long)newsegs, 0, 1);

	return 0;
}

int prepare_hugepage_range(unsigned long addr, unsigned long len)
{
	if (within_hugepage_high_range(addr, len))
		return 0;
	else if ((addr < 0x100000000UL) && ((addr+len) < 0x100000000UL)) {
		int err;
		/* Yes, we need both tests, in case addr+len overflows
		 * 64-bit arithmetic */
		err = open_low_hpage_segs(current->mm,
					  LOW_ESID_MASK(addr, len));
		if (err)
			printk(KERN_DEBUG "prepare_hugepage_range(%lx, %lx)"
			       " failed (segs: 0x%04hx)\n", addr, len,
			       LOW_ESID_MASK(addr, len));
		return err;
	}

	return -EINVAL;
}

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;

	return page;
}

int pmd_huge(pmd_t pmd)
{
	return 0;
}

struct page *
follow_huge_pmd(struct mm_struct *mm, unsigned long address,
		pmd_t *pmd, int write)
{
	BUG();
	return NULL;
}

/* Because we have an exclusive hugepage region which lies within the
 * normal user address space, we have to take special measures to make
 * non-huge mmap()s evade the hugepage reserved regions. */
unsigned long arch_get_unmapped_area(struct file *filp, unsigned long addr,
				     unsigned long len, unsigned long pgoff,
				     unsigned long flags)
{
	struct mm_struct *mm = current->mm;
	struct vm_area_struct *vma;
	unsigned long start_addr;

	if (len > TASK_SIZE)
		return -ENOMEM;

	if (addr) {
		addr = PAGE_ALIGN(addr);
		vma = find_vma(mm, addr);
		if (((TASK_SIZE - len) >= addr)
		    && (!vma || (addr+len) <= vma->vm_start)
		    && !is_hugepage_only_range(mm, addr,len))
			return addr;
	}
	start_addr = addr = mm->free_area_cache;

full_search:
	vma = find_vma(mm, addr);
	while (TASK_SIZE - len >= addr) {
		BUG_ON(vma && (addr >= vma->vm_end));

		if (touches_hugepage_low_range(mm, addr, len)) {
			addr = ALIGN(addr+1, 1<<SID_SHIFT);
			vma = find_vma(mm, addr);
			continue;
		}
		if (touches_hugepage_high_range(addr, len)) {
			addr = TASK_HPAGE_END;
			vma = find_vma(mm, addr);
			continue;
		}
		if (!vma || addr + len <= vma->vm_start) {
			/*
			 * Remember the place where we stopped the search:
			 */
			mm->free_area_cache = addr + len;
			return addr;
		}
		addr = vma->vm_end;
		vma = vma->vm_next;
	}

	/* Make sure we didn't miss any holes */
	if (start_addr != TASK_UNMAPPED_BASE) {
		start_addr = addr = TASK_UNMAPPED_BASE;
		goto full_search;
	}
	return -ENOMEM;
}

/*
 * This mmap-allocator allocates new areas top-down from below the
 * stack's low limit (the base):
 *
 * Because we have an exclusive hugepage region which lies within the
 * normal user address space, we have to take special measures to make
 * non-huge mmap()s evade the hugepage reserved regions.
 */
unsigned long
arch_get_unmapped_area_topdown(struct file *filp, const unsigned long addr0,
			  const unsigned long len, const unsigned long pgoff,
			  const unsigned long flags)
{
	struct vm_area_struct *vma, *prev_vma;
	struct mm_struct *mm = current->mm;
	unsigned long base = mm->mmap_base, addr = addr0;
	int first_time = 1;

	/* requested length too big for entire address space */
	if (len > TASK_SIZE)
		return -ENOMEM;

	/* dont allow allocations above current base */
	if (mm->free_area_cache > base)
		mm->free_area_cache = base;

	/* requesting a specific address */
	if (addr) {
		addr = PAGE_ALIGN(addr);
		vma = find_vma(mm, addr);
		if (TASK_SIZE - len >= addr &&
				(!vma || addr + len <= vma->vm_start)
				&& !is_hugepage_only_range(mm, addr,len))
			return addr;
	}

try_again:
	/* make sure it can fit in the remaining address space */
	if (mm->free_area_cache < len)
		goto fail;

	/* either no address requested or cant fit in requested address hole */
	addr = (mm->free_area_cache - len) & PAGE_MASK;
	do {
hugepage_recheck:
		if (touches_hugepage_low_range(mm, addr, len)) {
			addr = (addr & ((~0) << SID_SHIFT)) - len;
			goto hugepage_recheck;
		} else if (touches_hugepage_high_range(addr, len)) {
			addr = TASK_HPAGE_BASE - len;
		}

		/*
		 * Lookup failure means no vma is above this address,
		 * i.e. return with success:
		 */
 	 	if (!(vma = find_vma_prev(mm, addr, &prev_vma)))
			return addr;

		/*
		 * new region fits between prev_vma->vm_end and
		 * vma->vm_start, use it:
		 */
		if (addr+len <= vma->vm_start &&
				(!prev_vma || (addr >= prev_vma->vm_end)))
			/* remember the address as a hint for next time */
			return (mm->free_area_cache = addr);
		else
			/* pull free_area_cache down to the first hole */
			if (mm->free_area_cache == vma->vm_end)
				mm->free_area_cache = vma->vm_start;

		/* try just below the current vma->vm_start */
		addr = vma->vm_start-len;
	} while (len <= vma->vm_start);

fail:
	/*
	 * if hint left us with no space for the requested
	 * mapping then try again:
	 */
	if (first_time) {
		mm->free_area_cache = base;
		first_time = 0;
		goto try_again;
	}
	/*
	 * A failed mmap() very likely causes application failure,
	 * so fall back to the bottom-up function here. This scenario
	 * can happen with large stack limits and large mmap()
	 * allocations.
	 */
	mm->free_area_cache = TASK_UNMAPPED_BASE;
	addr = arch_get_unmapped_area(filp, addr0, len, pgoff, flags);
	/*
	 * Restore the topdown base:
	 */
	mm->free_area_cache = base;

	return addr;
}

static unsigned long htlb_get_low_area(unsigned long len, u16 segmask)
{
	unsigned long addr = 0;
	struct vm_area_struct *vma;

	vma = find_vma(current->mm, addr);
	while (addr + len <= 0x100000000UL) {
		BUG_ON(vma && (addr >= vma->vm_end)); /* invariant */

		if (! __within_hugepage_low_range(addr, len, segmask)) {
			addr = ALIGN(addr+1, 1<<SID_SHIFT);
			vma = find_vma(current->mm, addr);
			continue;
		}

		if (!vma || (addr + len) <= vma->vm_start)
			return addr;
		addr = ALIGN(vma->vm_end, HPAGE_SIZE);
		/* Depending on segmask this might not be a confirmed
		 * hugepage region, so the ALIGN could have skipped
		 * some VMAs */
		vma = find_vma(current->mm, addr);
	}

	return -ENOMEM;
}

static unsigned long htlb_get_high_area(unsigned long len)
{
	unsigned long addr = TASK_HPAGE_BASE;
	struct vm_area_struct *vma;

	vma = find_vma(current->mm, addr);
	for (vma = find_vma(current->mm, addr);
	     addr + len <= TASK_HPAGE_END;
	     vma = vma->vm_next) {
		BUG_ON(vma && (addr >= vma->vm_end)); /* invariant */
		BUG_ON(! within_hugepage_high_range(addr, len));

		if (!vma || (addr + len) <= vma->vm_start)
			return addr;
		addr = ALIGN(vma->vm_end, HPAGE_SIZE);
		/* Because we're in a hugepage region, this alignment
		 * should not skip us over any VMAs */
	}

	return -ENOMEM;
}

unsigned long hugetlb_get_unmapped_area(struct file *file, unsigned long addr,
					unsigned long len, unsigned long pgoff,
					unsigned long flags)
{
	if (len & ~HPAGE_MASK)
		return -EINVAL;

	if (!cpu_has_feature(CPU_FTR_16M_PAGE))
		return -EINVAL;

	if (test_thread_flag(TIF_32BIT)) {
		int lastshift = 0;
		u16 segmask, cursegs = current->mm->context.htlb_segs;

		/* First see if we can do the mapping in the existing
		 * low hpage segments */
		addr = htlb_get_low_area(len, cursegs);
		if (addr != -ENOMEM)
			return addr;

		for (segmask = LOW_ESID_MASK(0x100000000UL-len, len);
		     ! lastshift; segmask >>=1) {
			if (segmask & 1)
				lastshift = 1;

			addr = htlb_get_low_area(len, cursegs | segmask);
			if ((addr != -ENOMEM)
			    && open_low_hpage_segs(current->mm, segmask) == 0)
				return addr;
		}
		printk(KERN_DEBUG "hugetlb_get_unmapped_area() unable to open"
		       " enough segments\n");
		return -ENOMEM;
	} else {
		return htlb_get_high_area(len);
	}
}

void hugetlb_mm_free_pgd(struct mm_struct *mm)
{
	int i;
	pgd_t *pgdir;

	spin_lock(&mm->page_table_lock);

	pgdir = mm->context.huge_pgdir;
	if (! pgdir)
		goto out;

	mm->context.huge_pgdir = NULL;

	/* cleanup any hugepte pages leftover */
	for (i = 0; i < PTRS_PER_HUGEPGD; i++) {
		pud_t *pud = (pud_t *)(pgdir + i);

		if (! pud_none(*pud)) {
			pte_t *pte = (pte_t *)pud_page(*pud);
			struct page *ptepage = virt_to_page(pte);

			ptepage->mapping = NULL;

			BUG_ON(memcmp(pte, empty_zero_page, PAGE_SIZE));
			kmem_cache_free(zero_cache, pte);
		}
		pud_clear(pud);
	}

	BUG_ON(memcmp(pgdir, empty_zero_page, PAGE_SIZE));
	kmem_cache_free(zero_cache, pgdir);

 out:
	spin_unlock(&mm->page_table_lock);
}

int hash_huge_page(struct mm_struct *mm, unsigned long access,
		   unsigned long ea, unsigned long vsid, int local)
{
	pte_t *ptep;
	unsigned long va, vpn;
	pte_t old_pte, new_pte;
	unsigned long hpteflags, prpn;
	long slot;
	int err = 1;

	spin_lock(&mm->page_table_lock);

	ptep = huge_pte_offset(mm, ea);

	/* Search the Linux page table for a match with va */
	va = (vsid << 28) | (ea & 0x0fffffff);
	vpn = va >> HPAGE_SHIFT;

	/*
	 * If no pte found or not present, send the problem up to
	 * do_page_fault
	 */
	if (unlikely(!ptep || pte_none(*ptep)))
		goto out;

/* 	BUG_ON(pte_bad(*ptep)); */

	/* 
	 * Check the user's access rights to the page.  If access should be
	 * prevented then send the problem up to do_page_fault.
	 */
	if (unlikely(access & ~pte_val(*ptep)))
		goto out;
	/*
	 * At this point, we have a pte (old_pte) which can be used to build
	 * or update an HPTE. There are 2 cases:
	 *
	 * 1. There is a valid (present) pte with no associated HPTE (this is 
	 *	the most common case)
	 * 2. There is a valid (present) pte with an associated HPTE. The
	 *	current values of the pp bits in the HPTE prevent access
	 *	because we are doing software DIRTY bit management and the
	 *	page is currently not DIRTY. 
	 */


	old_pte = *ptep;
	new_pte = old_pte;

	hpteflags = 0x2 | (! (pte_val(new_pte) & _PAGE_RW));
 	/* _PAGE_EXEC -> HW_NO_EXEC since it's inverted */
	hpteflags |= ((pte_val(new_pte) & _PAGE_EXEC) ? 0 : HW_NO_EXEC);

	/* Check if pte already has an hpte (case 2) */
	if (unlikely(pte_val(old_pte) & _PAGE_HASHPTE)) {
		/* There MIGHT be an HPTE for this pte */
		unsigned long hash, slot;

		hash = hpt_hash(vpn, 1);
		if (pte_val(old_pte) & _PAGE_SECONDARY)
			hash = ~hash;
		slot = (hash & htab_hash_mask) * HPTES_PER_GROUP;
		slot += (pte_val(old_pte) & _PAGE_GROUP_IX) >> 12;

		if (ppc_md.hpte_updatepp(slot, hpteflags, va, 1, local) == -1)
			pte_val(old_pte) &= ~_PAGE_HPTEFLAGS;
	}

	if (likely(!(pte_val(old_pte) & _PAGE_HASHPTE))) {
		unsigned long hash = hpt_hash(vpn, 1);
		unsigned long hpte_group;

		prpn = pte_pfn(old_pte);

repeat:
		hpte_group = ((hash & htab_hash_mask) *
			      HPTES_PER_GROUP) & ~0x7UL;

		/* Update the linux pte with the HPTE slot */
		pte_val(new_pte) &= ~_PAGE_HPTEFLAGS;
		pte_val(new_pte) |= _PAGE_HASHPTE;

		/* Add in WIMG bits */
		/* XXX We should store these in the pte */
		hpteflags |= _PAGE_COHERENT;

		slot = ppc_md.hpte_insert(hpte_group, va, prpn, 0,
					  hpteflags, 0, 1);

		/* Primary is full, try the secondary */
		if (unlikely(slot == -1)) {
			pte_val(new_pte) |= _PAGE_SECONDARY;
			hpte_group = ((~hash & htab_hash_mask) *
				      HPTES_PER_GROUP) & ~0x7UL; 
			slot = ppc_md.hpte_insert(hpte_group, va, prpn,
						  1, hpteflags, 0, 1);
			if (slot == -1) {
				if (mftb() & 0x1)
					hpte_group = ((hash & htab_hash_mask) * HPTES_PER_GROUP) & ~0x7UL;

				ppc_md.hpte_remove(hpte_group);
				goto repeat;
                        }
		}

		if (unlikely(slot == -2))
			panic("hash_huge_page: pte_insert failed\n");

		pte_val(new_pte) |= (slot<<12) & _PAGE_GROUP_IX;

		/* 
		 * No need to use ldarx/stdcx here because all who
		 * might be updating the pte will hold the
		 * page_table_lock
		 */
		*ptep = new_pte;
	}

	err = 0;

 out:
	spin_unlock(&mm->page_table_lock);

	return err;
}