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
|
// SPDX-License-Identifier: GPL-2.0
/*
* Copyright IBM Corp. 2008
*
* Guest page hinting for unused pages.
*
* Author(s): Martin Schwidefsky <schwidefsky@de.ibm.com>
*/
#include <linux/kernel.h>
#include <linux/errno.h>
#include <linux/types.h>
#include <linux/mm.h>
#include <linux/memblock.h>
#include <linux/gfp.h>
#include <linux/init.h>
#include <asm/asm-extable.h>
#include <asm/facility.h>
#include <asm/page-states.h>
static int cmma_flag = 1;
static int __init cmma(char *str)
{
bool enabled;
if (!kstrtobool(str, &enabled))
cmma_flag = enabled;
return 1;
}
__setup("cmma=", cmma);
static inline int cmma_test_essa(void)
{
unsigned long tmp = 0;
int rc = -EOPNOTSUPP;
/* test ESSA_GET_STATE */
asm volatile(
" .insn rrf,0xb9ab0000,%[tmp],%[tmp],%[cmd],0\n"
"0: la %[rc],0\n"
"1:\n"
EX_TABLE(0b, 1b)
: [rc] "+&d" (rc), [tmp] "+&d" (tmp)
: [cmd] "i" (ESSA_GET_STATE));
return rc;
}
void __init cmma_init(void)
{
if (!cmma_flag)
return;
if (cmma_test_essa()) {
cmma_flag = 0;
return;
}
if (test_facility(147))
cmma_flag = 2;
}
static __always_inline void essa(unsigned long paddr, unsigned char cmd)
{
unsigned long rc;
asm volatile(
" .insn rrf,0xb9ab0000,%[rc],%[paddr],%[cmd],0"
: [rc] "=d" (rc)
: [paddr] "d" (paddr),
[cmd] "i" (cmd));
}
static __always_inline void __set_page_state(struct page *page, int order, unsigned char cmd)
{
unsigned long paddr = page_to_phys(page);
unsigned long num_pages = 1UL << order;
while (num_pages--) {
essa(paddr, cmd);
paddr += PAGE_SIZE;
}
}
static inline void set_page_unused(struct page *page, int order)
{
__set_page_state(page, order, ESSA_SET_UNUSED);
}
static inline void set_page_stable_dat(struct page *page, int order)
{
__set_page_state(page, order, ESSA_SET_STABLE);
}
static inline void set_page_stable_nodat(struct page *page, int order)
{
__set_page_state(page, order, ESSA_SET_STABLE_NODAT);
}
static void mark_kernel_pmd(pud_t *pud, unsigned long addr, unsigned long end)
{
unsigned long next;
struct page *page;
pmd_t *pmd;
pmd = pmd_offset(pud, addr);
do {
next = pmd_addr_end(addr, end);
if (pmd_none(*pmd) || pmd_large(*pmd))
continue;
page = phys_to_page(pmd_val(*pmd));
set_bit(PG_arch_1, &page->flags);
} while (pmd++, addr = next, addr != end);
}
static void mark_kernel_pud(p4d_t *p4d, unsigned long addr, unsigned long end)
{
unsigned long next;
struct page *page;
pud_t *pud;
int i;
pud = pud_offset(p4d, addr);
do {
next = pud_addr_end(addr, end);
if (pud_none(*pud) || pud_large(*pud))
continue;
if (!pud_folded(*pud)) {
page = phys_to_page(pud_val(*pud));
for (i = 0; i < 4; i++)
set_bit(PG_arch_1, &page[i].flags);
}
mark_kernel_pmd(pud, addr, next);
} while (pud++, addr = next, addr != end);
}
static void mark_kernel_p4d(pgd_t *pgd, unsigned long addr, unsigned long end)
{
unsigned long next;
struct page *page;
p4d_t *p4d;
int i;
p4d = p4d_offset(pgd, addr);
do {
next = p4d_addr_end(addr, end);
if (p4d_none(*p4d))
continue;
if (!p4d_folded(*p4d)) {
page = phys_to_page(p4d_val(*p4d));
for (i = 0; i < 4; i++)
set_bit(PG_arch_1, &page[i].flags);
}
mark_kernel_pud(p4d, addr, next);
} while (p4d++, addr = next, addr != end);
}
static void mark_kernel_pgd(void)
{
unsigned long addr, next, max_addr;
struct page *page;
pgd_t *pgd;
int i;
addr = 0;
/*
* Figure out maximum virtual address accessible with the
* kernel ASCE. This is required to keep the page table walker
* from accessing non-existent entries.
*/
max_addr = (S390_lowcore.kernel_asce.val & _ASCE_TYPE_MASK) >> 2;
max_addr = 1UL << (max_addr * 11 + 31);
pgd = pgd_offset_k(addr);
do {
next = pgd_addr_end(addr, max_addr);
if (pgd_none(*pgd))
continue;
if (!pgd_folded(*pgd)) {
page = phys_to_page(pgd_val(*pgd));
for (i = 0; i < 4; i++)
set_bit(PG_arch_1, &page[i].flags);
}
mark_kernel_p4d(pgd, addr, next);
} while (pgd++, addr = next, addr != max_addr);
}
void __init cmma_init_nodat(void)
{
struct page *page;
unsigned long start, end, ix;
int i;
if (cmma_flag < 2)
return;
/* Mark pages used in kernel page tables */
mark_kernel_pgd();
page = virt_to_page(&swapper_pg_dir);
for (i = 0; i < 4; i++)
set_bit(PG_arch_1, &page[i].flags);
page = virt_to_page(&invalid_pg_dir);
for (i = 0; i < 4; i++)
set_bit(PG_arch_1, &page[i].flags);
/* Set all kernel pages not used for page tables to stable/no-dat */
for_each_mem_pfn_range(i, MAX_NUMNODES, &start, &end, NULL) {
page = pfn_to_page(start);
for (ix = start; ix < end; ix++, page++) {
if (__test_and_clear_bit(PG_arch_1, &page->flags))
continue; /* skip page table pages */
if (!list_empty(&page->lru))
continue; /* skip free pages */
set_page_stable_nodat(page, 0);
}
}
}
void arch_free_page(struct page *page, int order)
{
if (!cmma_flag)
return;
set_page_unused(page, order);
}
void arch_alloc_page(struct page *page, int order)
{
if (!cmma_flag)
return;
if (cmma_flag < 2)
set_page_stable_dat(page, order);
else
set_page_stable_nodat(page, order);
}
void arch_set_page_dat(struct page *page, int order)
{
if (!cmma_flag)
return;
set_page_stable_dat(page, order);
}
|