memory.c 10.9 KB
Newer Older
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
/*
 * drivers/base/memory.c - basic Memory class support
 *
 * Written by Matt Tolentino <matthew.e.tolentino@intel.com>
 *            Dave Hansen <haveblue@us.ibm.com>
 *
 * This file provides the necessary infrastructure to represent
 * a SPARSEMEM-memory-model system's physical memory in /sysfs.
 * All arch-independent code that assumes MEMORY_HOTPLUG requires
 * SPARSEMEM should be contained here, or in mm/memory_hotplug.c.
 */

#include <linux/sysdev.h>
#include <linux/module.h>
#include <linux/init.h>
#include <linux/topology.h>
17
#include <linux/capability.h>
18
19
20
21
22
#include <linux/device.h>
#include <linux/memory.h>
#include <linux/kobject.h>
#include <linux/memory_hotplug.h>
#include <linux/mm.h>
23
#include <linux/mutex.h>
24
25
26
27
28
29
#include <asm/atomic.h>
#include <asm/uaccess.h>

#define MEMORY_CLASS_NAME	"memory"

static struct sysdev_class memory_sysdev_class = {
30
	.name = MEMORY_CLASS_NAME,
31
32
};

33
static const char *memory_uevent_name(struct kset *kset, struct kobject *kobj)
34
35
36
37
{
	return MEMORY_CLASS_NAME;
}

38
static int memory_uevent(struct kset *kset, struct kobject *obj, struct kobj_uevent_env *env)
39
40
41
42
43
44
{
	int retval = 0;

	return retval;
}

45
46
47
static struct kset_uevent_ops memory_uevent_ops = {
	.name		= memory_uevent_name,
	.uevent		= memory_uevent,
48
49
};

50
static BLOCKING_NOTIFIER_HEAD(memory_chain);
51

52
int register_memory_notifier(struct notifier_block *nb)
53
{
54
        return blocking_notifier_chain_register(&memory_chain, nb);
55
}
56
EXPORT_SYMBOL(register_memory_notifier);
57

58
void unregister_memory_notifier(struct notifier_block *nb)
59
{
60
        blocking_notifier_chain_unregister(&memory_chain, nb);
61
}
62
EXPORT_SYMBOL(unregister_memory_notifier);
63
64
65
66

/*
 * register_memory - Setup a sysfs device for a memory block
 */
67
68
static
int register_memory(struct memory_block *memory, struct mem_section *section)
69
70
71
72
73
74
75
76
77
78
79
{
	int error;

	memory->sysdev.cls = &memory_sysdev_class;
	memory->sysdev.id = __section_nr(section);

	error = sysdev_register(&memory->sysdev);
	return error;
}

static void
80
unregister_memory(struct memory_block *memory, struct mem_section *section)
81
82
83
84
{
	BUG_ON(memory->sysdev.cls != &memory_sysdev_class);
	BUG_ON(memory->sysdev.id != __section_nr(section));

85
86
	/* drop the ref. we got in remove_memory_block() */
	kobject_put(&memory->sysdev.kobj);
87
88
89
90
91
92
93
94
	sysdev_unregister(&memory->sysdev);
}

/*
 * use this as the physical section index that this memsection
 * uses.
 */

95
96
static ssize_t show_mem_phys_index(struct sys_device *dev,
			struct sysdev_attribute *attr, char *buf)
97
98
99
100
101
102
103
104
105
{
	struct memory_block *mem =
		container_of(dev, struct memory_block, sysdev);
	return sprintf(buf, "%08lx\n", mem->phys_index);
}

/*
 * online, offline, going offline, etc.
 */
106
107
static ssize_t show_mem_state(struct sys_device *dev,
			struct sysdev_attribute *attr, char *buf)
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
{
	struct memory_block *mem =
		container_of(dev, struct memory_block, sysdev);
	ssize_t len = 0;

	/*
	 * We can probably put these states in a nice little array
	 * so that they're not open-coded
	 */
	switch (mem->state) {
		case MEM_ONLINE:
			len = sprintf(buf, "online\n");
			break;
		case MEM_OFFLINE:
			len = sprintf(buf, "offline\n");
			break;
		case MEM_GOING_OFFLINE:
			len = sprintf(buf, "going-offline\n");
			break;
		default:
			len = sprintf(buf, "ERROR-UNKNOWN-%ld\n",
					mem->state);
			WARN_ON(1);
			break;
	}

	return len;
}

137
int memory_notify(unsigned long val, void *v)
138
{
139
	return blocking_notifier_call_chain(&memory_chain, val, v);
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
}

/*
 * MEMORY_HOTPLUG depends on SPARSEMEM in mm/Kconfig, so it is
 * OK to have direct references to sparsemem variables in here.
 */
static int
memory_block_action(struct memory_block *mem, unsigned long action)
{
	int i;
	unsigned long psection;
	unsigned long start_pfn, start_paddr;
	struct page *first_page;
	int ret;
	int old_state = mem->state;

	psection = mem->phys_index;
	first_page = pfn_to_page(psection << PFN_SECTION_SHIFT);

	/*
	 * The probe routines leave the pages reserved, just
	 * as the bootmem code does.  Make sure they're still
	 * that way.
	 */
	if (action == MEM_ONLINE) {
		for (i = 0; i < PAGES_PER_SECTION; i++) {
			if (PageReserved(first_page+i))
				continue;

			printk(KERN_WARNING "section number %ld page number %d "
				"not reserved, was it already online? \n",
				psection, i);
			return -EBUSY;
		}
	}

	switch (action) {
		case MEM_ONLINE:
			start_pfn = page_to_pfn(first_page);
			ret = online_pages(start_pfn, PAGES_PER_SECTION);
			break;
		case MEM_OFFLINE:
			mem->state = MEM_GOING_OFFLINE;
			start_paddr = page_to_pfn(first_page) << PAGE_SHIFT;
			ret = remove_memory(start_paddr,
					    PAGES_PER_SECTION << PAGE_SHIFT);
			if (ret) {
				mem->state = old_state;
				break;
			}
			break;
		default:
			printk(KERN_WARNING "%s(%p, %ld) unknown action: %ld\n",
193
					__func__, mem, action, action);
194
195
196
197
198
199
200
201
202
203
204
			WARN_ON(1);
			ret = -EINVAL;
	}

	return ret;
}

static int memory_block_change_state(struct memory_block *mem,
		unsigned long to_state, unsigned long from_state_req)
{
	int ret = 0;
205
	mutex_lock(&mem->state_mutex);
206
207
208
209
210
211
212
213
214
215
216

	if (mem->state != from_state_req) {
		ret = -EINVAL;
		goto out;
	}

	ret = memory_block_action(mem, to_state);
	if (!ret)
		mem->state = to_state;

out:
217
	mutex_unlock(&mem->state_mutex);
218
219
220
221
	return ret;
}

static ssize_t
222
223
store_mem_state(struct sys_device *dev,
		struct sysdev_attribute *attr, const char *buf, size_t count)
224
225
226
227
228
229
230
231
{
	struct memory_block *mem;
	unsigned int phys_section_nr;
	int ret = -EINVAL;

	mem = container_of(dev, struct memory_block, sysdev);
	phys_section_nr = mem->phys_index;

232
	if (!present_section_nr(phys_section_nr))
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
		goto out;

	if (!strncmp(buf, "online", min((int)count, 6)))
		ret = memory_block_change_state(mem, MEM_ONLINE, MEM_OFFLINE);
	else if(!strncmp(buf, "offline", min((int)count, 7)))
		ret = memory_block_change_state(mem, MEM_OFFLINE, MEM_ONLINE);
out:
	if (ret)
		return ret;
	return count;
}

/*
 * phys_device is a bad name for this.  What I really want
 * is a way to differentiate between memory ranges that
 * are part of physical devices that constitute
 * a complete removable unit or fru.
 * i.e. do these ranges belong to the same physical device,
 * s.t. if I offline all of these sections I can then
 * remove the physical device?
 */
254
255
static ssize_t show_phys_device(struct sys_device *dev,
				struct sysdev_attribute *attr, char *buf)
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
{
	struct memory_block *mem =
		container_of(dev, struct memory_block, sysdev);
	return sprintf(buf, "%d\n", mem->phys_device);
}

static SYSDEV_ATTR(phys_index, 0444, show_mem_phys_index, NULL);
static SYSDEV_ATTR(state, 0644, show_mem_state, store_mem_state);
static SYSDEV_ATTR(phys_device, 0444, show_phys_device, NULL);

#define mem_create_simple_file(mem, attr_name)	\
	sysdev_create_file(&mem->sysdev, &attr_##attr_name)
#define mem_remove_simple_file(mem, attr_name)	\
	sysdev_remove_file(&mem->sysdev, &attr_##attr_name)

/*
 * Block size attribute stuff
 */
static ssize_t
print_block_size(struct class *class, char *buf)
{
	return sprintf(buf, "%lx\n", (unsigned long)PAGES_PER_SECTION * PAGE_SIZE);
}

static CLASS_ATTR(block_size_bytes, 0444, print_block_size, NULL);

static int block_size_init(void)
{
284
285
	return sysfs_create_file(&memory_sysdev_class.kset.kobj,
				&class_attr_block_size_bytes.attr);
286
287
288
289
290
291
292
293
294
295
}

/*
 * Some architectures will have custom drivers to do this, and
 * will not need to do it from userspace.  The fake hot-add code
 * as well as ppc64 will do all of their discovery in userspace
 * and will require this interface.
 */
#ifdef CONFIG_ARCH_MEMORY_PROBE
static ssize_t
296
memory_probe_store(struct class *class, const char *buf, size_t count)
297
298
{
	u64 phys_addr;
299
	int nid;
300
301
302
303
	int ret;

	phys_addr = simple_strtoull(buf, NULL, 0);

304
305
	nid = memory_add_physaddr_to_nid(phys_addr);
	ret = add_memory(nid, phys_addr, PAGES_PER_SECTION << PAGE_SHIFT);
306
307
308
309
310
311
312
313
314
315

	if (ret)
		count = ret;

	return count;
}
static CLASS_ATTR(probe, 0700, NULL, memory_probe_store);

static int memory_probe_init(void)
{
316
317
	return sysfs_create_file(&memory_sysdev_class.kset.kobj,
				&class_attr_probe.attr);
318
319
}
#else
320
321
322
323
static inline int memory_probe_init(void)
{
	return 0;
}
324
325
326
327
328
329
330
331
332
333
334
#endif

/*
 * Note that phys_device is optional.  It is here to allow for
 * differentiation between which *physical* devices each
 * section belongs to...
 */

static int add_memory_block(unsigned long node_id, struct mem_section *section,
		     unsigned long state, int phys_device)
{
335
	struct memory_block *mem = kzalloc(sizeof(*mem), GFP_KERNEL);
336
337
338
339
340
341
342
	int ret = 0;

	if (!mem)
		return -ENOMEM;

	mem->phys_index = __section_nr(section);
	mem->state = state;
343
	mutex_init(&mem->state_mutex);
344
345
	mem->phys_device = phys_device;

346
	ret = register_memory(mem, section);
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
	if (!ret)
		ret = mem_create_simple_file(mem, phys_index);
	if (!ret)
		ret = mem_create_simple_file(mem, state);
	if (!ret)
		ret = mem_create_simple_file(mem, phys_device);

	return ret;
}

/*
 * For now, we have a linear search to go find the appropriate
 * memory_block corresponding to a particular phys_index. If
 * this gets to be a real problem, we can always use a radix
 * tree or something here.
 *
 * This could be made generic for all sysdev classes.
 */
static struct memory_block *find_memory_block(struct mem_section *section)
{
	struct kobject *kobj;
	struct sys_device *sysdev;
	struct memory_block *mem;
	char name[sizeof(MEMORY_CLASS_NAME) + 9 + 1];

	/*
	 * This only works because we know that section == sysdev->id
	 * slightly redundant with sysdev_register()
	 */
	sprintf(&name[0], "%s%d", MEMORY_CLASS_NAME, __section_nr(section));

	kobj = kset_find_obj(&memory_sysdev_class.kset, name);
	if (!kobj)
		return NULL;

	sysdev = container_of(kobj, struct sys_device, kobj);
	mem = container_of(sysdev, struct memory_block, sysdev);

	return mem;
}

int remove_memory_block(unsigned long node_id, struct mem_section *section,
		int phys_device)
{
	struct memory_block *mem;

	mem = find_memory_block(section);
	mem_remove_simple_file(mem, phys_index);
	mem_remove_simple_file(mem, state);
	mem_remove_simple_file(mem, phys_device);
397
	unregister_memory(mem, section);
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412

	return 0;
}

/*
 * need an interface for the VM to add new memory regions,
 * but without onlining it.
 */
int register_new_memory(struct mem_section *section)
{
	return add_memory_block(0, section, MEM_OFFLINE, 0);
}

int unregister_memory_section(struct mem_section *section)
{
413
	if (!present_section(section))
414
415
416
417
418
419
420
421
422
423
424
425
		return -EINVAL;

	return remove_memory_block(0, section, 0);
}

/*
 * Initialize the sysfs support for memory devices...
 */
int __init memory_dev_init(void)
{
	unsigned int i;
	int ret;
426
	int err;
427

428
	memory_sysdev_class.kset.uevent_ops = &memory_uevent_ops;
429
	ret = sysdev_class_register(&memory_sysdev_class);
430
431
	if (ret)
		goto out;
432
433
434
435
436
437

	/*
	 * Create entries for memory sections that were found
	 * during boot and have been initialized
	 */
	for (i = 0; i < NR_MEM_SECTIONS; i++) {
438
		if (!present_section_nr(i))
439
			continue;
440
441
442
		err = add_memory_block(0, __nr_to_section(i), MEM_ONLINE, 0);
		if (!ret)
			ret = err;
443
444
	}

445
446
447
448
449
450
451
452
	err = memory_probe_init();
	if (!ret)
		ret = err;
	err = block_size_init();
	if (!ret)
		ret = err;
out:
	if (ret)
453
		printk(KERN_ERR "%s() failed: %d\n", __func__, ret);
454
455
	return ret;
}