aboutsummaryrefslogtreecommitdiff
path: root/sys/amd64/amd64/minidump_machdep.c
blob: 1f7af43044ee6c7f1fba2cdc424172b25aa0469c (plain) (blame)
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
/*-
 * Copyright (c) 2006 Peter Wemm
 * All rights reserved.
 *
 * Redistribution and use in source and binary forms, with or without
 * modification, are permitted provided that the following conditions
 * are met:
 *
 * 1. Redistributions of source code must retain the above copyright
 *    notice, this list of conditions and the following disclaimer.
 * 2. Redistributions in binary form must reproduce the above copyright
 *    notice, this list of conditions and the following disclaimer in the
 *    documentation and/or other materials provided with the distribution.
 *
 * THIS SOFTWARE IS PROVIDED BY THE AUTHOR ``AS IS'' AND ANY EXPRESS OR
 * IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES
 * OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED.
 * IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR ANY DIRECT, INDIRECT,
 * INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT
 * NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
 * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
 * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
 * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF
 * THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
 */

#include <sys/cdefs.h>
__FBSDID("$FreeBSD$");

#include <sys/param.h>
#include <sys/systm.h>
#include <sys/conf.h>
#include <sys/cons.h>
#include <sys/kernel.h>
#include <sys/kerneldump.h>
#include <sys/msgbuf.h>
#include <vm/vm.h>
#include <vm/pmap.h>
#include <machine/atomic.h>
#include <machine/elf.h>
#include <machine/md_var.h>
#include <machine/vmparam.h>
#include <machine/minidump.h>

CTASSERT(sizeof(struct kerneldumpheader) == 512);

/*
 * Don't touch the first SIZEOF_METADATA bytes on the dump device. This
 * is to protect us from metadata and to protect metadata from us.
 */
#define	SIZEOF_METADATA		(64*1024)

#define	MD_ALIGN(x)	(((off_t)(x) + PAGE_MASK) & ~PAGE_MASK)
#define	DEV_ALIGN(x)	(((off_t)(x) + (DEV_BSIZE-1)) & ~(DEV_BSIZE-1))

extern uint64_t KPDPphys;

uint64_t *vm_page_dump;
int vm_page_dump_size;

static struct kerneldumpheader kdh;
static off_t dumplo;

/* Handle chunked writes. */
static size_t fragsz;
static void *dump_va;
static size_t counter, progress;

CTASSERT(sizeof(*vm_page_dump) == 8);

static int
is_dumpable(vm_paddr_t pa)
{
	int i;

	for (i = 0; dump_avail[i] != 0 || dump_avail[i + 1] != 0; i += 2) {
		if (pa >= dump_avail[i] && pa < dump_avail[i + 1])
			return (1);
	}
	return (0);
}

/* XXX should be MI */
static void
mkdumpheader(struct kerneldumpheader *kdh, uint32_t archver, uint64_t dumplen,
    uint32_t blksz)
{

	bzero(kdh, sizeof(*kdh));
	strncpy(kdh->magic, KERNELDUMPMAGIC, sizeof(kdh->magic));
	strncpy(kdh->architecture, MACHINE_ARCH, sizeof(kdh->architecture));
	kdh->version = htod32(KERNELDUMPVERSION);
	kdh->architectureversion = htod32(archver);
	kdh->dumplength = htod64(dumplen);
	kdh->dumptime = htod64(time_second);
	kdh->blocksize = htod32(blksz);
	strncpy(kdh->hostname, hostname, sizeof(kdh->hostname));
	strncpy(kdh->versionstring, version, sizeof(kdh->versionstring));
	if (panicstr != NULL)
		strncpy(kdh->panicstring, panicstr, sizeof(kdh->panicstring));
	kdh->parity = kerneldump_parity(kdh);
}

#define PG2MB(pgs) (((pgs) + (1 << 8) - 1) >> 8)

static int
blk_flush(struct dumperinfo *di)
{
	int error;

	if (fragsz == 0)
		return (0);

	error = di->dumper(di->priv, dump_va, 0, dumplo, fragsz);
	dumplo += fragsz;
	fragsz = 0;
	return (error);
}

static int
blk_write(struct dumperinfo *di, char *ptr, vm_paddr_t pa, size_t sz)
{
	size_t len;
	int error, i, c;

	error = 0;
	if ((sz % PAGE_SIZE) != 0) {
		printf("size not page aligned\n");
		return (EINVAL);
	}
	if (ptr != NULL && pa != 0) {
		printf("cant have both va and pa!\n");
		return (EINVAL);
	}
	if (pa != 0 && (((uintptr_t)ptr) % PAGE_SIZE) != 0) {
		printf("address not page aligned\n");
		return (EINVAL);
	}
	if (ptr != NULL) {
		/* If we're doing a virtual dump, flush any pre-existing pa pages */
		error = blk_flush(di);
		if (error)
			return (error);
	}
	while (sz) {
		len = (MAXDUMPPGS * PAGE_SIZE) - fragsz;
		if (len > sz)
			len = sz;
		counter += len;
		progress -= len;
		if (counter >> 24) {
			printf(" %ld", PG2MB(progress >> PAGE_SHIFT));
			counter &= (1<<24) - 1;
		}
		if (ptr) {
			error = di->dumper(di->priv, ptr, 0, dumplo, len);
			if (error)
				return (error);
			dumplo += len;
			ptr += len;
			sz -= len;
		} else {
			for (i = 0; i < len; i += PAGE_SIZE)
				dump_va = pmap_kenter_temporary(pa + i, (i + fragsz) >> PAGE_SHIFT);
			fragsz += len;
			pa += len;
			sz -= len;
			if (fragsz == (MAXDUMPPGS * PAGE_SIZE)) {
				error = blk_flush(di);
				if (error)
					return (error);
			}
		}

		/* Check for user abort. */
		c = cncheckc();
		if (c == 0x03)
			return (ECANCELED);
		if (c != -1)
			printf(" (CTRL-C to abort) ");
	}

	return (0);
}

/* A fake page table page, to avoid having to handle both 4K and 2M pages */
static pt_entry_t fakept[NPTEPG];

void
minidumpsys(struct dumperinfo *di)
{
	uint64_t dumpsize;
	uint32_t ptesize;
	vm_offset_t va;
	int error;
	uint64_t bits;
	uint64_t *pdp, *pd, *pt, pa;
	int i, j, k, bit;
	struct minidumphdr mdhdr;

	counter = 0;
	/* Walk page table pages, set bits in vm_page_dump */
	ptesize = 0;
	pdp = (uint64_t *)PHYS_TO_DMAP(KPDPphys);
	for (va = KERNBASE; va < kernel_vm_end; va += NBPDR) {
		i = (va >> PDPSHIFT) & ((1ul << NPDPEPGSHIFT) - 1);
		/*
		 * We always write a page, even if it is zero. Each
		 * page written corresponds to 2MB of space
		 */
		ptesize += PAGE_SIZE;
		if ((pdp[i] & PG_V) == 0)
			continue;
		pd = (uint64_t *)PHYS_TO_DMAP(pdp[i] & PG_FRAME);
		j = ((va >> PDRSHIFT) & ((1ul << NPDEPGSHIFT) - 1));
		if ((pd[j] & (PG_PS | PG_V)) == (PG_PS | PG_V))  {
			/* This is an entire 2M page. */
			pa = pd[j] & PG_FRAME & ~PDRMASK;
			for (k = 0; k < NPTEPG; k++) {
				if (is_dumpable(pa))
					dump_add_page(pa);
				pa += PAGE_SIZE;
			}
			continue;
		}
		if ((pd[j] & PG_V) == PG_V) {
			/* set bit for each valid page in this 2MB block */
			pt = (uint64_t *)PHYS_TO_DMAP(pd[j] & PG_FRAME);
			for (k = 0; k < NPTEPG; k++) {
				if ((pt[k] & PG_V) == PG_V) {
					pa = pt[k] & PG_FRAME;
					if (is_dumpable(pa))
						dump_add_page(pa);
				}
			}
		} else {
			/* nothing, we're going to dump a null page */
		}
	}

	/* Calculate dump size. */
	dumpsize = ptesize;
	dumpsize += round_page(msgbufp->msg_size);
	dumpsize += round_page(vm_page_dump_size);
	for (i = 0; i < vm_page_dump_size / sizeof(*vm_page_dump); i++) {
		bits = vm_page_dump[i];
		while (bits) {
			bit = bsfq(bits);
			pa = (((uint64_t)i * sizeof(*vm_page_dump) * NBBY) + bit) * PAGE_SIZE;
			/* Clear out undumpable pages now if needed */
			if (is_dumpable(pa)) {
				dumpsize += PAGE_SIZE;
			} else {
				dump_drop_page(pa);
			}
			bits &= ~(1ul << bit);
		}
	}
	dumpsize += PAGE_SIZE;

	/* Determine dump offset on device. */
	if (di->mediasize < SIZEOF_METADATA + dumpsize + sizeof(kdh) * 2) {
		error = ENOSPC;
		goto fail;
	}
	dumplo = di->mediaoffset + di->mediasize - dumpsize;
	dumplo -= sizeof(kdh) * 2;
	progress = dumpsize;

	/* Initialize mdhdr */
	bzero(&mdhdr, sizeof(mdhdr));
	strcpy(mdhdr.magic, MINIDUMP_MAGIC);
	mdhdr.version = MINIDUMP_VERSION;
	mdhdr.msgbufsize = msgbufp->msg_size;
	mdhdr.bitmapsize = vm_page_dump_size;
	mdhdr.ptesize = ptesize;
	mdhdr.kernbase = KERNBASE;
	mdhdr.dmapbase = DMAP_MIN_ADDRESS;
	mdhdr.dmapend = DMAP_MAX_ADDRESS;

	mkdumpheader(&kdh, KERNELDUMP_AMD64_VERSION, dumpsize, di->blocksize);

	printf("Physical memory: %ju MB\n", ptoa((uintmax_t)physmem) / 1048576);
	printf("Dumping %llu MB:", (long long)dumpsize >> 20);

	/* Dump leader */
	error = di->dumper(di->priv, &kdh, 0, dumplo, sizeof(kdh));
	if (error)
		goto fail;
	dumplo += sizeof(kdh);

	/* Dump my header */
	bzero(&fakept, sizeof(fakept));
	bcopy(&mdhdr, &fakept, sizeof(mdhdr));
	error = blk_write(di, (char *)&fakept, 0, PAGE_SIZE);
	if (error)
		goto fail;

	/* Dump msgbuf up front */
	error = blk_write(di, (char *)msgbufp->msg_ptr, 0, round_page(msgbufp->msg_size));
	if (error)
		goto fail;

	/* Dump bitmap */
	error = blk_write(di, (char *)vm_page_dump, 0, round_page(vm_page_dump_size));
	if (error)
		goto fail;

	/* Dump kernel page table pages */
	pdp = (uint64_t *)PHYS_TO_DMAP(KPDPphys);
	for (va = KERNBASE; va < kernel_vm_end; va += NBPDR) {
		i = (va >> PDPSHIFT) & ((1ul << NPDPEPGSHIFT) - 1);
		/* We always write a page, even if it is zero */
		if ((pdp[i] & PG_V) == 0) {
			bzero(fakept, sizeof(fakept));
			error = blk_write(di, (char *)&fakept, 0, PAGE_SIZE);
			if (error)
				goto fail;
			/* flush, in case we reuse fakept in the same block */
			error = blk_flush(di);
			if (error)
				goto fail;
			continue;
		}
		pd = (uint64_t *)PHYS_TO_DMAP(pdp[i] & PG_FRAME);
		j = ((va >> PDRSHIFT) & ((1ul << NPDEPGSHIFT) - 1));
		if ((pd[j] & (PG_PS | PG_V)) == (PG_PS | PG_V))  {
			/* This is a single 2M block. Generate a fake PTP */
			pa = pd[j] & PG_FRAME & ~PDRMASK;
			for (k = 0; k < NPTEPG; k++) {
				fakept[k] = (pa + (k * PAGE_SIZE)) | PG_V | PG_RW | PG_A | PG_M;
			}
			error = blk_write(di, (char *)&fakept, 0, PAGE_SIZE);
			if (error)
				goto fail;
			/* flush, in case we reuse fakept in the same block */
			error = blk_flush(di);
			if (error)
				goto fail;
			continue;
		}
		if ((pd[j] & PG_V) == PG_V) {
			pt = (uint64_t *)PHYS_TO_DMAP(pd[j] & PG_FRAME);
			error = blk_write(di, (char *)pt, 0, PAGE_SIZE);
			if (error)
				goto fail;
		} else {
			bzero(fakept, sizeof(fakept));
			error = blk_write(di, (char *)&fakept, 0, PAGE_SIZE);
			if (error)
				goto fail;
			/* flush, in case we reuse fakept in the same block */
			error = blk_flush(di);
			if (error)
				goto fail;
		}
	}

	/* Dump memory chunks */
	/* XXX cluster it up and use blk_dump() */
	for (i = 0; i < vm_page_dump_size / sizeof(*vm_page_dump); i++) {
		bits = vm_page_dump[i];
		while (bits) {
			bit = bsfq(bits);
			pa = (((uint64_t)i * sizeof(*vm_page_dump) * NBBY) + bit) * PAGE_SIZE;
			error = blk_write(di, 0, pa, PAGE_SIZE);
			if (error)
				goto fail;
			bits &= ~(1ul << bit);
		}
	}

	error = blk_flush(di);
	if (error)
		goto fail;

	/* Dump trailer */
	error = di->dumper(di->priv, &kdh, 0, dumplo, sizeof(kdh));
	if (error)
		goto fail;
	dumplo += sizeof(kdh);

	/* Signal completion, signoff and exit stage left. */
	di->dumper(di->priv, NULL, 0, 0, 0);
	printf("\nDump complete\n");
	return;

 fail:
	if (error < 0)
		error = -error;

	if (error == ECANCELED)
		printf("\nDump aborted\n");
	else if (error == ENOSPC)
		printf("\nDump failed. Partition too small.\n");
	else
		printf("\n** DUMP FAILED (ERROR %d) **\n", error);
}

void
dump_add_page(vm_paddr_t pa)
{
	int idx, bit;

	pa >>= PAGE_SHIFT;
	idx = pa >> 6;		/* 2^6 = 64 */
	bit = pa & 63;
	atomic_set_long(&vm_page_dump[idx], 1ul << bit);
}

void
dump_drop_page(vm_paddr_t pa)
{
	int idx, bit;

	pa >>= PAGE_SHIFT;
	idx = pa >> 6;		/* 2^6 = 64 */
	bit = pa & 63;
	atomic_clear_long(&vm_page_dump[idx], 1ul << bit);
}