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
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
|
/*P:700 The pagetable code, on the other hand, still shows the scars of
* previous encounters. It's functional, and as neat as it can be in the
* circumstances, but be wary, for these things are subtle and break easily.
* The Guest provides a virtual to physical mapping, but we can neither trust
* it nor use it: we verify and convert it here then point the CPU to the
* converted Guest pages when running the Guest. :*/
/* Copyright (C) Rusty Russell IBM Corporation 2006.
* GPL v2 and any later version */
#include <linux/mm.h>
#include <linux/types.h>
#include <linux/spinlock.h>
#include <linux/random.h>
#include <linux/percpu.h>
#include <asm/tlbflush.h>
#include <asm/uaccess.h>
#include "lg.h"
/*M:008 We hold reference to pages, which prevents them from being swapped.
* It'd be nice to have a callback in the "struct mm_struct" when Linux wants
* to swap out. If we had this, and a shrinker callback to trim PTE pages, we
* could probably consider launching Guests as non-root. :*/
/*H:300
* The Page Table Code
*
* We use two-level page tables for the Guest. If you're not entirely
* comfortable with virtual addresses, physical addresses and page tables then
* I recommend you review arch/x86/lguest/boot.c's "Page Table Handling" (with
* diagrams!).
*
* The Guest keeps page tables, but we maintain the actual ones here: these are
* called "shadow" page tables. Which is a very Guest-centric name: these are
* the real page tables the CPU uses, although we keep them up to date to
* reflect the Guest's. (See what I mean about weird naming? Since when do
* shadows reflect anything?)
*
* Anyway, this is the most complicated part of the Host code. There are seven
* parts to this:
* (i) Looking up a page table entry when the Guest faults,
* (ii) Making sure the Guest stack is mapped,
* (iii) Setting up a page table entry when the Guest tells us one has changed,
* (iv) Switching page tables,
* (v) Flushing (throwing away) page tables,
* (vi) Mapping the Switcher when the Guest is about to run,
* (vii) Setting up the page tables initially.
:*/
/* 1024 entries in a page table page maps 1024 pages: 4MB. The Switcher is
* conveniently placed at the top 4MB, so it uses a separate, complete PTE
* page. */
#define SWITCHER_PGD_INDEX (PTRS_PER_PGD - 1)
/* We actually need a separate PTE page for each CPU. Remember that after the
* Switcher code itself comes two pages for each CPU, and we don't want this
* CPU's guest to see the pages of any other CPU. */
static DEFINE_PER_CPU(pte_t *, switcher_pte_pages);
#define switcher_pte_page(cpu) per_cpu(switcher_pte_pages, cpu)
/*H:320 The page table code is curly enough to need helper functions to keep it
* clear and clean.
*
* There are two functions which return pointers to the shadow (aka "real")
* page tables.
*
* spgd_addr() takes the virtual address and returns a pointer to the top-level
* page directory entry (PGD) for that address. Since we keep track of several
* page tables, the "i" argument tells us which one we're interested in (it's
* usually the current one). */
static pgd_t *spgd_addr(struct lg_cpu *cpu, u32 i, unsigned long vaddr)
{
unsigned int index = pgd_index(vaddr);
/* We kill any Guest trying to touch the Switcher addresses. */
if (index >= SWITCHER_PGD_INDEX) {
kill_guest(cpu, "attempt to access switcher pages");
index = 0;
}
/* Return a pointer index'th pgd entry for the i'th page table. */
return &cpu->lg->pgdirs[i].pgdir[index];
}
/* This routine then takes the page directory entry returned above, which
* contains the address of the page table entry (PTE) page. It then returns a
* pointer to the PTE entry for the given address. */
static pte_t *spte_addr(pgd_t spgd, unsigned long vaddr)
{
pte_t *page = __va(pgd_pfn(spgd) << PAGE_SHIFT);
/* You should never call this if the PGD entry wasn't valid */
BUG_ON(!(pgd_flags(spgd) & _PAGE_PRESENT));
return &page[(vaddr >> PAGE_SHIFT) % PTRS_PER_PTE];
}
/* These two functions just like the above two, except they access the Guest
* page tables. Hence they return a Guest address. */
static unsigned long gpgd_addr(struct lg_cpu *cpu, unsigned long vaddr)
{
unsigned int index = vaddr >> (PGDIR_SHIFT);
return cpu->lg->pgdirs[cpu->cpu_pgd].gpgdir + index * sizeof(pgd_t);
}
static unsigned long gpte_addr(pgd_t gpgd, unsigned long vaddr)
{
unsigned long gpage = pgd_pfn(gpgd) << PAGE_SHIFT;
BUG_ON(!(pgd_flags(gpgd) & _PAGE_PRESENT));
return gpage + ((vaddr>>PAGE_SHIFT) % PTRS_PER_PTE) * sizeof(pte_t);
}
/*:*/
/*M:014 get_pfn is slow; it takes the mmap sem and calls get_user_pages. We
* could probably try to grab batches of pages here as an optimization
* (ie. pre-faulting). :*/
/*H:350 This routine takes a page number given by the Guest and converts it to
* an actual, physical page number. It can fail for several reasons: the
* virtual address might not be mapped by the Launcher, the write flag is set
* and the page is read-only, or the write flag was set and the page was
* shared so had to be copied, but we ran out of memory.
*
* This holds a reference to the page, so release_pte() is careful to put that
* back. */
static unsigned long get_pfn(unsigned long virtpfn, int write)
{
struct page *page;
/* This value indicates failure. */
unsigned long ret = -1UL;
/* get_user_pages() is a complex interface: it gets the "struct
* vm_area_struct" and "struct page" assocated with a range of pages.
* It also needs the task's mmap_sem held, and is not very quick.
* It returns the number of pages it got. */
down_read(¤t->mm->mmap_sem);
if (get_user_pages(current, current->mm, virtpfn << PAGE_SHIFT,
1, write, 1, &page, NULL) == 1)
ret = page_to_pfn(page);
up_read(¤t->mm->mmap_sem);
return ret;
}
/*H:340 Converting a Guest page table entry to a shadow (ie. real) page table
* entry can be a little tricky. The flags are (almost) the same, but the
* Guest PTE contains a virtual page number: the CPU needs the real page
* number. */
static pte_t gpte_to_spte(struct lg_cpu *cpu, pte_t gpte, int write)
{
unsigned long pfn, base, flags;
/* The Guest sets the global flag, because it thinks that it is using
* PGE. We only told it to use PGE so it would tell us whether it was
* flushing a kernel mapping or a userspace mapping. We don't actually
* use the global bit, so throw it away. */
flags = (pte_flags(gpte) & ~_PAGE_GLOBAL);
/* The Guest's pages are offset inside the Launcher. */
base = (unsigned long)cpu->lg->mem_base / PAGE_SIZE;
/* We need a temporary "unsigned long" variable to hold the answer from
* get_pfn(), because it returns 0xFFFFFFFF on failure, which wouldn't
* fit in spte.pfn. get_pfn() finds the real physical number of the
* page, given the virtual number. */
pfn = get_pfn(base + pte_pfn(gpte), write);
if (pfn == -1UL) {
kill_guest(cpu, "failed to get page %lu", pte_pfn(gpte));
/* When we destroy the Guest, we'll go through the shadow page
* tables and release_pte() them. Make sure we don't think
* this one is valid! */
flags = 0;
}
/* Now we assemble our shadow PTE from the page number and flags. */
return pfn_pte(pfn, __pgprot(flags));
}
/*H:460 And to complete the chain, release_pte() looks like this: */
static void release_pte(pte_t pte)
{
/* Remember that get_user_pages() took a reference to the page, in
* get_pfn()? We have to put it back now. */
if (pte_flags(pte) & _PAGE_PRESENT)
put_page(pfn_to_page(pte_pfn(pte)));
}
/*:*/
static void check_gpte(struct lg_cpu *cpu, pte_t gpte)
{
if ((pte_flags(gpte) & _PAGE_PSE) ||
pte_pfn(gpte) >= cpu->lg->pfn_limit)
kill_guest(cpu, "bad page table entry");
}
static void check_gpgd(struct lg_cpu *cpu, pgd_t gpgd)
{
if ((pgd_flags(gpgd) & ~_PAGE_TABLE) ||
(pgd_pfn(gpgd) >= cpu->lg->pfn_limit))
kill_guest(cpu, "bad page directory entry");
}
/*H:330
* (i) Looking up a page table entry when the Guest faults.
*
* We saw this call in run_guest(): when we see a page fault in the Guest, we
* come here. That's because we only set up the shadow page tables lazily as
* they're needed, so we get page faults all the time and quietly fix them up
* and return to the Guest without it knowing.
*
* If we fixed up the fault (ie. we mapped the address), this routine returns
* true. Otherwise, it was a real fault and we need to tell the Guest. */
int demand_page(struct lg_cpu *cpu, unsigned long vaddr, int errcode)
{
pgd_t gpgd;
pgd_t *spgd;
unsigned long gpte_ptr;
pte_t gpte;
pte_t *spte;
/* First step: get the top-level Guest page table entry. */
gpgd = lgread(cpu, gpgd_addr(cpu, vaddr), pgd_t);
/* Toplevel not present? We can't map it in. */
if (!(pgd_flags(gpgd) & _PAGE_PRESENT))
return 0;
/* Now look at the matching shadow entry. */
spgd = spgd_addr(cpu, cpu->cpu_pgd,
|