summaryrefslogtreecommitdiffstats
path: root/arch/powerpc/mm/book3s64/hash_tlb.c
blob: 21fcad97ae80dd46fe28b6f195150d82f52979d8 (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
// SPDX-License-Identifier: GPL-2.0-or-later
/*
 * This file contains the routines for flushing entries from the
 * TLB and MMU hash table.
 *
 *  Derived from arch/ppc64/mm/init.c:
 *    Copyright (C) 1995-1996 Gary Thomas (gdt@linuxppc.org)
 *
 *  Modifications by Paul Mackerras (PowerMac) (paulus@cs.anu.edu.au)
 *  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.
 */

#include <linux/kernel.h>
#include <linux/mm.h>
#include <linux/percpu.h>
#include <linux/hardirq.h>
#include <asm/tlbflush.h>
#include <asm/tlb.h>
#include <asm/bug.h>
#include <asm/pte-walk.h>


#include <trace/events/thp.h>

DEFINE_PER_CPU(struct ppc64_tlb_batch, ppc64_tlb_batch);

/*
 * A linux PTE was changed and the corresponding hash table entry
 * neesd to be flushed. This function will either perform the flush
 * immediately or will batch it up if the current CPU has an active
 * batch on it.
 */
void hpte_need_flush(struct mm_struct *mm, unsigned long addr,
		     pte_t *ptep, unsigned long pte, int huge)
{
	unsigned long vpn;
	struct ppc64_tlb_batch *batch = &get_cpu_var(ppc64_tlb_batch);
	unsigned long vsid;
	unsigned int psize;
	int ssize;
	real_pte_t rpte;
	int i, offset;

	i = batch->index;

	/*
	 * Get page size (maybe move back to caller).
	 *
	 * NOTE: when using special 64K mappings in 4K environment like
	 * for SPEs, we obtain the page size from the slice, which thus
	 * must still exist (and thus the VMA not reused) at the time
	 * of this call
	 */
	if (huge) {
#ifdef CONFIG_HUGETLB_PAGE
		psize = get_slice_psize(mm, addr);
		/* Mask the address for the correct page size */
		addr &= ~((1UL << mmu_psize_defs[psize].shift) - 1);
		if (unlikely(psize == MMU_PAGE_16G))
			offset = PTRS_PER_PUD;
		else
			offset = PTRS_PER_PMD;
#else
		BUG();
		psize = pte_pagesize_index(mm, addr, pte); /* shutup gcc */
#endif
	} else {
		psize = pte_pagesize_index(mm, addr, pte);
		/*
		 * Mask the address for the standard page size.  If we
		 * have a 64k page kernel, but the hardware does not
		 * support 64k pages, this might be different from the
		 * hardware page size encoded in the slice table.
		 */
		addr &= PAGE_MASK;
		offset = PTRS_PER_PTE;
	}


	/* Build full vaddr */
	if (!is_kernel_addr(addr)) {
		ssize = user_segment_size(addr);
		vsid = get_user_vsid(&mm->context, addr, ssize);
	} else {
		vsid = get_kernel_vsid(addr, mmu_kernel_ssize);
		ssize = mmu_kernel_ssize;
	}
	WARN_ON(vsid == 0);
	vpn = hpt_vpn(addr, vsid, ssize);
	rpte = __real_pte(__pte(pte), ptep, offset);

	/*
	 * Check if we have an active batch on this CPU. If not, just
	 * flush now and return.
	 */
	if (!batch->active) {
		flush_hash_page(vpn, rpte, psize, ssize, mm_is_thread_local(mm));
		put_cpu_var(ppc64_tlb_batch);
		return;
	}

	/*
	 * This can happen when we are in the middle of a TLB batch and
	 * we encounter memory pressure (eg copy_page_range when it tries
	 * to allocate a new pte). If we have to reclaim memory and end
	 * up scanning and resetting referenced bits then our batch context
	 * will change mid stream.
	 *
	 * We also need to ensure only one page size is present in a given
	 * batch
	 */
	if (i != 0 && (mm != batch->mm || batch->psize != psize ||
		       batch->ssize != ssize)) {
		__flush_tlb_pending(batch);
		i = 0;
	}
	if (i == 0) {
		batch->mm = mm;
		batch->psize = psize;
		batch->ssize = ssize;
	}
	batch->pte[i] = rpte;
	batch->vpn[i] = vpn;
	batch->index = ++i;
	if (i >= PPC64_TLB_BATCH_NR)
		__flush_tlb_pending(batch);
	put_cpu_var(ppc64_tlb_batch);
}

/*
 * This function is called when terminating an mmu batch or when a batch
 * is full. It will perform the flush of all the entries currently stored
 * in a batch.
 *
 * Must be called from within some kind of spinlock/non-preempt region...
 */
void __flush_tlb_pending(struct ppc64_tlb_batch *batch)
{
	int i, local;

	i = batch->index;
	local = mm_is_thread_local(batch->mm);
	if (i == 1)
		flush_hash_page(batch->vpn[0], batch->pte[0],
				batch->psize, batch->ssize, local);
	else
		flush_hash_range(i, local);
	batch->index = 0;
}

void hash__tlb_flush(struct mmu_gather *tlb)
{
	struct ppc64_tlb_batch *tlbbatch = &get_cpu_var(ppc64_tlb_batch);

	/*
	 * If there's a TLB batch pending, then we must flush it because the
	 * pages are going to be freed and we really don't want to have a CPU
	 * access a freed page because it has a stale TLB
	 */
	if (tlbbatch->index)
		__flush_tlb_pending(tlbbatch);

	put_cpu_var(ppc64_tlb_batch);
}

/**
 * __flush_hash_table_range - Flush all HPTEs for a given address range
 *                            from the hash table (and the TLB). But keeps
 *                            the linux PTEs intact.
 *
 * @start	: starting address
 * @end         : ending address (not included in the flush)
 *
 * This function is mostly to be used by some IO hotplug code in order
 * to remove all hash entries from a given address range used to map IO
 * space on a removed PCI-PCI bidge without tearing down the full mapping
 * since 64K pages may overlap with other bridges when using 64K pages
 * with 4K HW pages on IO space.
 *
 * Because of that usage pattern, it is implemented for small size rather
 * than speed.
 */
void __flush_hash_table_range(unsigned long start, unsigned long end)
{
	int hugepage_shift;
	unsigned long flags;

	start = ALIGN_DOWN(start, PAGE_SIZE);
	end = ALIGN(end, PAGE_SIZE);


	/*
	 * Note: Normally, we should only ever use a batch within a
	 * PTE locked section. This violates the rule, but will work
	 * since we don't actually modify the PTEs, we just flush the
	 * hash while leaving the PTEs intact (including their reference
	 * to being hashed). This is not the most performance oriented
	 * way to do things but is fine for our needs here.
	 */
	local_irq_save(flags);
	arch_enter_lazy_mmu_mode();
	for (; start < end; start += PAGE_SIZE) {
		pte_t *ptep = find_init_mm_pte(start, &hugepage_shift);
		unsigned long pte;

		if (ptep == NULL)
			continue;
		pte = pte_val(*ptep);
		if (!(pte & H_PAGE_HASHPTE))
			continue;
		hpte_need_flush(&init_mm, start, ptep, pte, hugepage_shift);
	}
	arch_leave_lazy_mmu_mode();
	local_irq_restore(flags);
}

void flush_hash_table_pmd_range(struct mm_struct *mm, pmd_t *pmd, unsigned long addr)
{
	pte_t *pte;
	pte_t *start_pte;
	unsigned long flags;

	addr = ALIGN_DOWN(addr, PMD_SIZE);
	/*
	 * Note: Normally, we should only ever use a batch within a
	 * PTE locked section. This violates the rule, but will work
	 * since we don't actually modify the PTEs, we just flush the
	 * hash while leaving the PTEs intact (including their reference
	 * to being hashed). This is not the most performance oriented
	 * way to do things but is fine for our needs here.
	 */
	local_irq_save(flags);
	arch_enter_lazy_mmu_mode();
	start_pte = pte_offset_map(pmd, addr);
	if (!start_pte)
		goto out;
	for (pte = start_pte; pte < start_pte + PTRS_PER_PTE; pte++) {
		unsigned long pteval = pte_val(*pte);
		if (pteval & H_PAGE_HASHPTE)
			hpte_need_flush(mm, addr, pte, pteval, 0);
		addr += PAGE_SIZE;
	}
	pte_unmap(start_pte);
out:
	arch_leave_lazy_mmu_mode();
	local_irq_restore(flags);
}