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
|
// SPDX-License-Identifier: GPL-2.0-or-later
/*
* Copyright (c) 2021-2024 Oracle. All Rights Reserved.
* Author: Darrick J. Wong <djwong@kernel.org>
*/
#include "xfs.h"
#include "xfs_fs.h"
#include "xfs_shared.h"
#include "xfs_format.h"
#include "xfs_trans_resv.h"
#include "xfs_mount.h"
#include "xfs_log_format.h"
#include "xfs_trans.h"
#include "xfs_inode.h"
#include "xfs_icache.h"
#include "xfs_bmap_util.h"
#include "xfs_iwalk.h"
#include "xfs_ialloc.h"
#include "xfs_sb.h"
#include "scrub/scrub.h"
#include "scrub/common.h"
#include "scrub/repair.h"
#include "scrub/xfile.h"
#include "scrub/xfarray.h"
#include "scrub/iscan.h"
#include "scrub/nlinks.h"
#include "scrub/trace.h"
/*
* Live Inode Link Count Repair
* ============================
*
* Use the live inode link count information that we collected to replace the
* nlink values of the incore inodes. A scrub->repair cycle should have left
* the live data and hooks active, so this is safe so long as we make sure the
* inode is locked.
*/
/*
* Correct the link count of the given inode. Because we have to grab locks
* and resources in a certain order, it's possible that this will be a no-op.
*/
STATIC int
xrep_nlinks_repair_inode(
struct xchk_nlink_ctrs *xnc)
{
struct xchk_nlink obs;
struct xfs_scrub *sc = xnc->sc;
struct xfs_mount *mp = sc->mp;
struct xfs_inode *ip = sc->ip;
uint64_t total_links;
uint64_t actual_nlink;
bool dirty = false;
int error;
xchk_ilock(sc, XFS_IOLOCK_EXCL);
error = xfs_trans_alloc(mp, &M_RES(mp)->tr_link, 0, 0, 0, &sc->tp);
if (error)
return error;
xchk_ilock(sc, XFS_ILOCK_EXCL);
xfs_trans_ijoin(sc->tp, ip, 0);
mutex_lock(&xnc->lock);
if (xchk_iscan_aborted(&xnc->collect_iscan)) {
error = -ECANCELED;
goto out_scanlock;
}
error = xfarray_load_sparse(xnc->nlinks, ip->i_ino, &obs);
if (error)
goto out_scanlock;
/*
* We're done accessing the shared scan data, so we can drop the lock.
* We still hold @ip's ILOCK, so its link count cannot change.
*/
mutex_unlock(&xnc->lock);
total_links = xchk_nlink_total(ip, &obs);
actual_nlink = VFS_I(ip)->i_nlink;
/*
* Non-directories cannot have directories pointing up to them.
*
* We previously set error to zero, but set it again because one static
* checker author fears that programmers will fail to maintain this
* invariant and built their tool to flag this as a security risk. A
* different tool author made their bot complain about the redundant
* store. This is a never-ending and stupid battle; both tools missed
* *actual bugs* elsewhere; and I no longer care.
*/
if (!S_ISDIR(VFS_I(ip)->i_mode) && obs.children != 0) {
trace_xrep_nlinks_unfixable_inode(mp, ip, &obs);
error = 0;
goto out_trans;
}
/*
* We did not find any links to this inode. If the inode agrees, we
* have nothing further to do. If not, the inode has a nonzero link
* count and we don't have anywhere to graft the child onto. Dropping
* a live inode's link count to zero can cause unexpected shutdowns in
* inactivation, so leave it alone.
*/
if (total_links == 0) {
if (actual_nlink != 0)
trace_xrep_nlinks_unfixable_inode(mp, ip, &obs);
goto out_trans;
}
/* Commit the new link count if it changed. */
if (total_links != actual_nlink) {
if (total_links > XFS_MAXLINK) {
trace_xrep_nlinks_unfixable_inode(mp, ip, &obs);
goto out_trans;
}
trace_xrep_nlinks_update_inode(mp, ip, &obs);
set_nlink(VFS_I(ip), total_links);
dirty = true;
}
if (!dirty) {
error = 0;
goto out_trans;
}
xfs_trans_log_inode(sc->tp, ip, XFS_ILOG_CORE);
error = xrep_trans_commit(sc);
xchk_iunlock(sc, XFS_ILOCK_EXCL | XFS_IOLOCK_EXCL);
return error;
out_scanlock:
mutex_unlock(&xnc->lock);
out_trans:
xchk_trans_cancel(sc);
xchk_iunlock(sc, XFS_ILOCK_EXCL | XFS_IOLOCK_EXCL);
return error;
}
/*
* Try to visit every inode in the filesystem for repairs. Move on if we can't
* grab an inode, since we're still making forward progress.
*/
static int
xrep_nlinks_iter(
struct xchk_nlink_ctrs *xnc,
struct xfs_inode **ipp)
{
int error;
do {
error = xchk_iscan_iter(&xnc->compare_iscan, ipp);
} while (error == -EBUSY);
return error;
}
/* Commit the new inode link counters. */
int
xrep_nlinks(
struct xfs_scrub *sc)
{
struct xchk_nlink_ctrs *xnc = sc->buf;
int error;
/*
* We need ftype for an accurate count of the number of child
* subdirectory links. Child subdirectories with a back link (dotdot
* entry) but no forward link are unfixable, so we cannot repair the
* link count of the parent directory based on the back link count
* alone. Filesystems without ftype support are rare (old V4) so we
* just skip out here.
*/
if (!xfs_has_ftype(sc->mp))
return -EOPNOTSUPP;
/*
* Use the inobt to walk all allocated inodes to compare and fix the
* link counts. Retry iget every tenth of a second for up to 30
* seconds -- even if repair misses a few inodes, we still try to fix
* as many of them as we can.
*/
xchk_iscan_start(sc, 30000, 100, &xnc->compare_iscan);
ASSERT(sc->ip == NULL);
while ((error = xrep_nlinks_iter(xnc, &sc->ip)) == 1) {
/*
* Commit the scrub transaction so that we can create repair
* transactions with the correct reservations.
*/
xchk_trans_cancel(sc);
error = xrep_nlinks_repair_inode(xnc);
xchk_iscan_mark_visited(&xnc->compare_iscan, sc->ip);
xchk_irele(sc, sc->ip);
sc->ip = NULL;
if (error)
break;
if (xchk_should_terminate(sc, &error))
break;
/*
* Create a new empty transaction so that we can advance the
* iscan cursor without deadlocking if the inobt has a cycle.
* We can only push the inactivation workqueues with an empty
* transaction.
*/
error = xchk_trans_alloc_empty(sc);
if (error)
break;
}
xchk_iscan_iter_finish(&xnc->compare_iscan);
xchk_iscan_teardown(&xnc->compare_iscan);
return error;
}
|
