linux-stable/fs/xfs/libxfs/xfs_dir2.c

771 lines
17 KiB
C
Raw Normal View History

// SPDX-License-Identifier: GPL-2.0
/*
* Copyright (c) 2000-2001,2005 Silicon Graphics, Inc.
* All Rights Reserved.
*/
#include "xfs.h"
#include "xfs_fs.h"
#include "xfs_shared.h"
#include "xfs_format.h"
#include "xfs_log_format.h"
#include "xfs_trans_resv.h"
#include "xfs_mount.h"
#include "xfs_inode.h"
#include "xfs_trans.h"
#include "xfs_bmap.h"
#include "xfs_dir2.h"
#include "xfs_dir2_priv.h"
#include "xfs_errortag.h"
#include "xfs_error.h"
xfs: event tracing support Convert the old xfs tracing support that could only be used with the out of tree kdb and xfsidbg patches to use the generic event tracer. To use it make sure CONFIG_EVENT_TRACING is enabled and then enable all xfs trace channels by: echo 1 > /sys/kernel/debug/tracing/events/xfs/enable or alternatively enable single events by just doing the same in one event subdirectory, e.g. echo 1 > /sys/kernel/debug/tracing/events/xfs/xfs_ihold/enable or set more complex filters, etc. In Documentation/trace/events.txt all this is desctribed in more detail. To reads the events do a cat /sys/kernel/debug/tracing/trace Compared to the last posting this patch converts the tracing mostly to the one tracepoint per callsite model that other users of the new tracing facility also employ. This allows a very fine-grained control of the tracing, a cleaner output of the traces and also enables the perf tool to use each tracepoint as a virtual performance counter, allowing us to e.g. count how often certain workloads git various spots in XFS. Take a look at http://lwn.net/Articles/346470/ for some examples. Also the btree tracing isn't included at all yet, as it will require additional core tracing features not in mainline yet, I plan to deliver it later. And the really nice thing about this patch is that it actually removes many lines of code while adding this nice functionality: fs/xfs/Makefile | 8 fs/xfs/linux-2.6/xfs_acl.c | 1 fs/xfs/linux-2.6/xfs_aops.c | 52 - fs/xfs/linux-2.6/xfs_aops.h | 2 fs/xfs/linux-2.6/xfs_buf.c | 117 +-- fs/xfs/linux-2.6/xfs_buf.h | 33 fs/xfs/linux-2.6/xfs_fs_subr.c | 3 fs/xfs/linux-2.6/xfs_ioctl.c | 1 fs/xfs/linux-2.6/xfs_ioctl32.c | 1 fs/xfs/linux-2.6/xfs_iops.c | 1 fs/xfs/linux-2.6/xfs_linux.h | 1 fs/xfs/linux-2.6/xfs_lrw.c | 87 -- fs/xfs/linux-2.6/xfs_lrw.h | 45 - fs/xfs/linux-2.6/xfs_super.c | 104 --- fs/xfs/linux-2.6/xfs_super.h | 7 fs/xfs/linux-2.6/xfs_sync.c | 1 fs/xfs/linux-2.6/xfs_trace.c | 75 ++ fs/xfs/linux-2.6/xfs_trace.h | 1369 +++++++++++++++++++++++++++++++++++++++++ fs/xfs/linux-2.6/xfs_vnode.h | 4 fs/xfs/quota/xfs_dquot.c | 110 --- fs/xfs/quota/xfs_dquot.h | 21 fs/xfs/quota/xfs_qm.c | 40 - fs/xfs/quota/xfs_qm_syscalls.c | 4 fs/xfs/support/ktrace.c | 323 --------- fs/xfs/support/ktrace.h | 85 -- fs/xfs/xfs.h | 16 fs/xfs/xfs_ag.h | 14 fs/xfs/xfs_alloc.c | 230 +----- fs/xfs/xfs_alloc.h | 27 fs/xfs/xfs_alloc_btree.c | 1 fs/xfs/xfs_attr.c | 107 --- fs/xfs/xfs_attr.h | 10 fs/xfs/xfs_attr_leaf.c | 14 fs/xfs/xfs_attr_sf.h | 40 - fs/xfs/xfs_bmap.c | 507 +++------------ fs/xfs/xfs_bmap.h | 49 - fs/xfs/xfs_bmap_btree.c | 6 fs/xfs/xfs_btree.c | 5 fs/xfs/xfs_btree_trace.h | 17 fs/xfs/xfs_buf_item.c | 87 -- fs/xfs/xfs_buf_item.h | 20 fs/xfs/xfs_da_btree.c | 3 fs/xfs/xfs_da_btree.h | 7 fs/xfs/xfs_dfrag.c | 2 fs/xfs/xfs_dir2.c | 8 fs/xfs/xfs_dir2_block.c | 20 fs/xfs/xfs_dir2_leaf.c | 21 fs/xfs/xfs_dir2_node.c | 27 fs/xfs/xfs_dir2_sf.c | 26 fs/xfs/xfs_dir2_trace.c | 216 ------ fs/xfs/xfs_dir2_trace.h | 72 -- fs/xfs/xfs_filestream.c | 8 fs/xfs/xfs_fsops.c | 2 fs/xfs/xfs_iget.c | 111 --- fs/xfs/xfs_inode.c | 67 -- fs/xfs/xfs_inode.h | 76 -- fs/xfs/xfs_inode_item.c | 5 fs/xfs/xfs_iomap.c | 85 -- fs/xfs/xfs_iomap.h | 8 fs/xfs/xfs_log.c | 181 +---- fs/xfs/xfs_log_priv.h | 20 fs/xfs/xfs_log_recover.c | 1 fs/xfs/xfs_mount.c | 2 fs/xfs/xfs_quota.h | 8 fs/xfs/xfs_rename.c | 1 fs/xfs/xfs_rtalloc.c | 1 fs/xfs/xfs_rw.c | 3 fs/xfs/xfs_trans.h | 47 + fs/xfs/xfs_trans_buf.c | 62 - fs/xfs/xfs_vnodeops.c | 8 70 files changed, 2151 insertions(+), 2592 deletions(-) Signed-off-by: Christoph Hellwig <hch@lst.de> Signed-off-by: Alex Elder <aelder@sgi.com>
2009-12-14 23:14:59 +00:00
#include "xfs_trace.h"
const struct xfs_name xfs_name_dotdot = {
.name = (const unsigned char *)"..",
.len = 2,
.type = XFS_DIR3_FT_DIR,
};
xfs: Add read-only support for dirent filetype field Add support for the file type field in directory entries so that readdir can return the type of the inode the dirent points to to userspace without first having to read the inode off disk. The encoding of the type field is a single byte that is added to the end of the directory entry name length. For all intents and purposes, it appends a "hidden" byte to the name field which contains the type information. As the directory entry is already of dynamic size, helpers are already required to access and decode the direct entry structures. Hence the relevent extraction and iteration helpers are updated to understand the hidden byte. Helpers for reading and writing the filetype field from the directory entries are also added. Only the read helpers are used by this patch. It also adds all the code necessary to read the type information out of the dirents on disk. Further we add the superblock feature bit and helpers to indicate that we understand the on-disk format change. This is not a compatible change - existing kernels cannot read the new format successfully - so an incompatible feature flag is added. We don't yet allow filesystems to mount with this flag yet - that will be added once write support is added. Finally, the code to take the type from the VFS, convert it to an XFS on-disk type and put it into the xfs_name structures passed around is added, but the directory code does not use this field yet. That will be in the next patch. Signed-off-by: Dave Chinner <dchinner@redhat.com> Reviewed-by: Mark Tinguely <tinguely@sgi.com> Signed-off-by: Ben Myers <bpm@sgi.com>
2013-08-12 10:50:09 +00:00
/*
* Convert inode mode to directory entry filetype
*/
unsigned char
xfs_mode_to_ftype(
int mode)
{
switch (mode & S_IFMT) {
case S_IFREG:
return XFS_DIR3_FT_REG_FILE;
case S_IFDIR:
return XFS_DIR3_FT_DIR;
case S_IFCHR:
return XFS_DIR3_FT_CHRDEV;
case S_IFBLK:
return XFS_DIR3_FT_BLKDEV;
case S_IFIFO:
return XFS_DIR3_FT_FIFO;
case S_IFSOCK:
return XFS_DIR3_FT_SOCK;
case S_IFLNK:
return XFS_DIR3_FT_SYMLINK;
default:
return XFS_DIR3_FT_UNKNOWN;
}
}
/*
* ASCII case-insensitive (ie. A-Z) support for directories that was
* used in IRIX.
*/
xfs_dahash_t
xfs_ascii_ci_hashname(
const struct xfs_name *name)
{
xfs_dahash_t hash;
int i;
for (i = 0, hash = 0; i < name->len; i++)
xfs: stabilize the dirent name transformation function used for ascii-ci dir hash computation Back in the old days, the "ascii-ci" feature was created to implement case-insensitive directory entry lookups for latin1-encoded names and remove the large overhead of Samba's case-insensitive lookup code. UTF8 names were not allowed, but nobody explicitly wrote in the documentation that this was only expected to work if the system used latin1 names. The kernel tolower function was selected to prepare names for hashed lookups. There's a major discrepancy in the function that computes directory entry hashes for filesystems that have ASCII case-insensitive lookups enabled. The root of this is that the kernel and glibc's tolower implementations have differing behavior for extended ASCII accented characters. I wrote a program to spit out characters for which the tolower() return value is different from the input: glibc tolower: 65:A 66:B 67:C 68:D 69:E 70:F 71:G 72:H 73:I 74:J 75:K 76:L 77:M 78:N 79:O 80:P 81:Q 82:R 83:S 84:T 85:U 86:V 87:W 88:X 89:Y 90:Z kernel tolower: 65:A 66:B 67:C 68:D 69:E 70:F 71:G 72:H 73:I 74:J 75:K 76:L 77:M 78:N 79:O 80:P 81:Q 82:R 83:S 84:T 85:U 86:V 87:W 88:X 89:Y 90:Z 192:À 193:Á 194:Â 195:Ã 196:Ä 197:Å 198:Æ 199:Ç 200:È 201:É 202:Ê 203:Ë 204:Ì 205:Í 206:Î 207:Ï 208:Ð 209:Ñ 210:Ò 211:Ó 212:Ô 213:Õ 214:Ö 215:× 216:Ø 217:Ù 218:Ú 219:Û 220:Ü 221:Ý 222:Þ Which means that the kernel and userspace do not agree on the hash value for a directory filename that contains those higher values. The hash values are written into the leaf index block of directories that are larger than two blocks in size, which means that xfs_repair will flag these directories as having corrupted hash indexes and rewrite the index with hash values that the kernel now will not recognize. Because the ascii-ci feature is not frequently enabled and the kernel touches filesystems far more frequently than xfs_repair does, fix this by encoding the kernel's toupper predicate and tolower functions into libxfs. Give the new functions less provocative names to make it really obvious that this is a pre-hash name preparation function, and nothing else. This change makes userspace's behavior consistent with the kernel. Found by auditing obfuscate_name in xfs_metadump as part of working on parent pointers, wondering how it could possibly work correctly with ci filesystems, writing a test tool to create a directory with hash-colliding names, and watching xfs_repair flag it. Signed-off-by: Darrick J. Wong <djwong@kernel.org> Reviewed-by: Christoph Hellwig <hch@lst.de>
2023-04-12 02:05:04 +00:00
hash = xfs_ascii_ci_xfrm(name->name[i]) ^ rol32(hash, 7);
return hash;
}
enum xfs_dacmp
xfs_ascii_ci_compname(
struct xfs_da_args *args,
const unsigned char *name,
int len)
{
enum xfs_dacmp result;
int i;
if (args->namelen != len)
return XFS_CMP_DIFFERENT;
result = XFS_CMP_EXACT;
for (i = 0; i < len; i++) {
if (args->name[i] == name[i])
continue;
xfs: stabilize the dirent name transformation function used for ascii-ci dir hash computation Back in the old days, the "ascii-ci" feature was created to implement case-insensitive directory entry lookups for latin1-encoded names and remove the large overhead of Samba's case-insensitive lookup code. UTF8 names were not allowed, but nobody explicitly wrote in the documentation that this was only expected to work if the system used latin1 names. The kernel tolower function was selected to prepare names for hashed lookups. There's a major discrepancy in the function that computes directory entry hashes for filesystems that have ASCII case-insensitive lookups enabled. The root of this is that the kernel and glibc's tolower implementations have differing behavior for extended ASCII accented characters. I wrote a program to spit out characters for which the tolower() return value is different from the input: glibc tolower: 65:A 66:B 67:C 68:D 69:E 70:F 71:G 72:H 73:I 74:J 75:K 76:L 77:M 78:N 79:O 80:P 81:Q 82:R 83:S 84:T 85:U 86:V 87:W 88:X 89:Y 90:Z kernel tolower: 65:A 66:B 67:C 68:D 69:E 70:F 71:G 72:H 73:I 74:J 75:K 76:L 77:M 78:N 79:O 80:P 81:Q 82:R 83:S 84:T 85:U 86:V 87:W 88:X 89:Y 90:Z 192:À 193:Á 194:Â 195:Ã 196:Ä 197:Å 198:Æ 199:Ç 200:È 201:É 202:Ê 203:Ë 204:Ì 205:Í 206:Î 207:Ï 208:Ð 209:Ñ 210:Ò 211:Ó 212:Ô 213:Õ 214:Ö 215:× 216:Ø 217:Ù 218:Ú 219:Û 220:Ü 221:Ý 222:Þ Which means that the kernel and userspace do not agree on the hash value for a directory filename that contains those higher values. The hash values are written into the leaf index block of directories that are larger than two blocks in size, which means that xfs_repair will flag these directories as having corrupted hash indexes and rewrite the index with hash values that the kernel now will not recognize. Because the ascii-ci feature is not frequently enabled and the kernel touches filesystems far more frequently than xfs_repair does, fix this by encoding the kernel's toupper predicate and tolower functions into libxfs. Give the new functions less provocative names to make it really obvious that this is a pre-hash name preparation function, and nothing else. This change makes userspace's behavior consistent with the kernel. Found by auditing obfuscate_name in xfs_metadump as part of working on parent pointers, wondering how it could possibly work correctly with ci filesystems, writing a test tool to create a directory with hash-colliding names, and watching xfs_repair flag it. Signed-off-by: Darrick J. Wong <djwong@kernel.org> Reviewed-by: Christoph Hellwig <hch@lst.de>
2023-04-12 02:05:04 +00:00
if (xfs_ascii_ci_xfrm(args->name[i]) !=
xfs_ascii_ci_xfrm(name[i]))
return XFS_CMP_DIFFERENT;
result = XFS_CMP_CASE;
}
return result;
}
int
xfs_da_mount(
struct xfs_mount *mp)
{
struct xfs_da_geometry *dageo;
ASSERT(mp->m_sb.sb_versionnum & XFS_SB_VERSION_DIRV2BIT);
ASSERT(xfs_dir2_dirblock_bytes(&mp->m_sb) <= XFS_MAX_BLOCKSIZE);
mp->m_dir_geo = kmem_zalloc(sizeof(struct xfs_da_geometry),
KM_MAYFAIL);
mp->m_attr_geo = kmem_zalloc(sizeof(struct xfs_da_geometry),
KM_MAYFAIL);
if (!mp->m_dir_geo || !mp->m_attr_geo) {
kmem_free(mp->m_dir_geo);
kmem_free(mp->m_attr_geo);
return -ENOMEM;
}
/* set up directory geometry */
dageo = mp->m_dir_geo;
dageo->blklog = mp->m_sb.sb_blocklog + mp->m_sb.sb_dirblklog;
dageo->fsblog = mp->m_sb.sb_blocklog;
dageo->blksize = xfs_dir2_dirblock_bytes(&mp->m_sb);
dageo->fsbcount = 1 << mp->m_sb.sb_dirblklog;
if (xfs_has_crc(mp)) {
dageo->node_hdr_size = sizeof(struct xfs_da3_node_hdr);
dageo->leaf_hdr_size = sizeof(struct xfs_dir3_leaf_hdr);
dageo->free_hdr_size = sizeof(struct xfs_dir3_free_hdr);
dageo->data_entry_offset =
sizeof(struct xfs_dir3_data_hdr);
} else {
dageo->node_hdr_size = sizeof(struct xfs_da_node_hdr);
dageo->leaf_hdr_size = sizeof(struct xfs_dir2_leaf_hdr);
dageo->free_hdr_size = sizeof(struct xfs_dir2_free_hdr);
dageo->data_entry_offset =
sizeof(struct xfs_dir2_data_hdr);
}
dageo->leaf_max_ents = (dageo->blksize - dageo->leaf_hdr_size) /
sizeof(struct xfs_dir2_leaf_entry);
dageo->free_max_bests = (dageo->blksize - dageo->free_hdr_size) /
sizeof(xfs_dir2_data_off_t);
dageo->data_first_offset = dageo->data_entry_offset +
xfs_dir2_data_entsize(mp, 1) +
xfs_dir2_data_entsize(mp, 2);
/*
* Now we've set up the block conversion variables, we can calculate the
* segment block constants using the geometry structure.
*/
dageo->datablk = xfs_dir2_byte_to_da(dageo, XFS_DIR2_DATA_OFFSET);
dageo->leafblk = xfs_dir2_byte_to_da(dageo, XFS_DIR2_LEAF_OFFSET);
dageo->freeblk = xfs_dir2_byte_to_da(dageo, XFS_DIR2_FREE_OFFSET);
dageo->node_ents = (dageo->blksize - dageo->node_hdr_size) /
(uint)sizeof(xfs_da_node_entry_t);
dageo->max_extents = (XFS_DIR2_MAX_SPACES * XFS_DIR2_SPACE_SIZE) >>
mp->m_sb.sb_blocklog;
dageo->magicpct = (dageo->blksize * 37) / 100;
/* set up attribute geometry - single fsb only */
dageo = mp->m_attr_geo;
dageo->blklog = mp->m_sb.sb_blocklog;
dageo->fsblog = mp->m_sb.sb_blocklog;
dageo->blksize = 1 << dageo->blklog;
dageo->fsbcount = 1;
dageo->node_hdr_size = mp->m_dir_geo->node_hdr_size;
dageo->node_ents = (dageo->blksize - dageo->node_hdr_size) /
(uint)sizeof(xfs_da_node_entry_t);
if (xfs_has_large_extent_counts(mp))
dageo->max_extents = XFS_MAX_EXTCNT_ATTR_FORK_LARGE;
else
dageo->max_extents = XFS_MAX_EXTCNT_ATTR_FORK_SMALL;
dageo->magicpct = (dageo->blksize * 37) / 100;
return 0;
}
void
xfs_da_unmount(
struct xfs_mount *mp)
{
kmem_free(mp->m_dir_geo);
kmem_free(mp->m_attr_geo);
}
/*
* Return 1 if directory contains only "." and "..".
*/
int
xfs_dir_isempty(
xfs_inode_t *dp)
{
xfs_dir2_sf_hdr_t *sfp;
ASSERT(S_ISDIR(VFS_I(dp)->i_mode));
if (dp->i_disk_size == 0) /* might happen during shutdown. */
return 1;
if (dp->i_disk_size > xfs_inode_data_fork_size(dp))
return 0;
sfp = (xfs_dir2_sf_hdr_t *)dp->i_df.if_u1.if_data;
return !sfp->count;
}
/*
* Validate a given inode number.
*/
int
xfs_dir_ino_validate(
xfs_mount_t *mp,
xfs_ino_t ino)
{
bool ino_ok = xfs_verify_dir_ino(mp, ino);
if (XFS_IS_CORRUPT(mp, !ino_ok) ||
XFS_TEST_ERROR(false, mp, XFS_ERRTAG_DIR_INO_VALIDATE)) {
xfs_warn(mp, "Invalid inode number 0x%Lx",
(unsigned long long) ino);
return -EFSCORRUPTED;
}
return 0;
}
/*
* Initialize a directory with its "." and ".." entries.
*/
int
xfs_dir_init(
xfs_trans_t *tp,
xfs_inode_t *dp,
xfs_inode_t *pdp)
{
struct xfs_da_args *args;
int error;
ASSERT(S_ISDIR(VFS_I(dp)->i_mode));
error = xfs_dir_ino_validate(tp->t_mountp, pdp->i_ino);
if (error)
return error;
args = kmem_zalloc(sizeof(*args), KM_NOFS);
if (!args)
return -ENOMEM;
args->geo = dp->i_mount->m_dir_geo;
args->dp = dp;
args->trans = tp;
error = xfs_dir2_sf_create(args, pdp->i_ino);
kmem_free(args);
return error;
}
/*
* Enter a name in a directory, or check for available space.
* If inum is 0, only the available space test is performed.
*/
int
xfs_dir_createname(
struct xfs_trans *tp,
struct xfs_inode *dp,
const struct xfs_name *name,
xfs_ino_t inum, /* new entry inode number */
xfs_extlen_t total) /* bmap's total block count */
{
struct xfs_da_args *args;
int rval;
bool v;
ASSERT(S_ISDIR(VFS_I(dp)->i_mode));
if (inum) {
rval = xfs_dir_ino_validate(tp->t_mountp, inum);
if (rval)
return rval;
XFS_STATS_INC(dp->i_mount, xs_dir_create);
}
args = kmem_zalloc(sizeof(*args), KM_NOFS);
if (!args)
return -ENOMEM;
args->geo = dp->i_mount->m_dir_geo;
args->name = name->name;
args->namelen = name->len;
args->filetype = name->type;
args->hashval = xfs_dir2_hashname(dp->i_mount, name);
args->inumber = inum;
args->dp = dp;
args->total = total;
args->whichfork = XFS_DATA_FORK;
args->trans = tp;
args->op_flags = XFS_DA_OP_ADDNAME | XFS_DA_OP_OKNOENT;
if (!inum)
args->op_flags |= XFS_DA_OP_JUSTCHECK;
if (dp->i_df.if_format == XFS_DINODE_FMT_LOCAL) {
rval = xfs_dir2_sf_addname(args);
goto out_free;
}
rval = xfs_dir2_isblock(args, &v);
if (rval)
goto out_free;
if (v) {
rval = xfs_dir2_block_addname(args);
goto out_free;
}
rval = xfs_dir2_isleaf(args, &v);
if (rval)
goto out_free;
if (v)
rval = xfs_dir2_leaf_addname(args);
else
rval = xfs_dir2_node_addname(args);
out_free:
kmem_free(args);
return rval;
}
/*
* If doing a CI lookup and case-insensitive match, dup actual name into
* args.value. Return EEXIST for success (ie. name found) or an error.
*/
int
xfs_dir_cilookup_result(
struct xfs_da_args *args,
const unsigned char *name,
int len)
{
if (args->cmpresult == XFS_CMP_DIFFERENT)
return -ENOENT;
if (args->cmpresult != XFS_CMP_CASE ||
!(args->op_flags & XFS_DA_OP_CILOOKUP))
return -EEXIST;
args->value = kmem_alloc(len, KM_NOFS | KM_MAYFAIL);
if (!args->value)
return -ENOMEM;
memcpy(args->value, name, len);
args->valuelen = len;
return -EEXIST;
}
/*
* Lookup a name in a directory, give back the inode number.
* If ci_name is not NULL, returns the actual name in ci_name if it differs
* to name, or ci_name->name is set to NULL for an exact match.
*/
int
xfs_dir_lookup(
struct xfs_trans *tp,
struct xfs_inode *dp,
const struct xfs_name *name,
xfs_ino_t *inum, /* out: inode number */
struct xfs_name *ci_name) /* out: actual name if CI match */
{
struct xfs_da_args *args;
int rval;
bool v;
int lock_mode;
ASSERT(S_ISDIR(VFS_I(dp)->i_mode));
XFS_STATS_INC(dp->i_mount, xs_dir_lookup);
/*
* We need to use KM_NOFS here so that lockdep will not throw false
* positive deadlock warnings on a non-transactional lookup path. It is
* safe to recurse into inode recalim in that case, but lockdep can't
* easily be taught about it. Hence KM_NOFS avoids having to add more
* lockdep Doing this avoids having to add a bunch of lockdep class
* annotations into the reclaim path for the ilock.
*/
args = kmem_zalloc(sizeof(*args), KM_NOFS);
args->geo = dp->i_mount->m_dir_geo;
args->name = name->name;
args->namelen = name->len;
args->filetype = name->type;
args->hashval = xfs_dir2_hashname(dp->i_mount, name);
args->dp = dp;
args->whichfork = XFS_DATA_FORK;
args->trans = tp;
args->op_flags = XFS_DA_OP_OKNOENT;
if (ci_name)
args->op_flags |= XFS_DA_OP_CILOOKUP;
xfs: stop holding ILOCK over filldir callbacks The recent change to the readdir locking made in 40194ec ("xfs: reinstate the ilock in xfs_readdir") for CXFS directory sanity was probably the wrong thing to do. Deep in the readdir code we can take page faults in the filldir callback, and so taking a page fault while holding an inode ilock creates a new set of locking issues that lockdep warns all over the place about. The locking order for regular inodes w.r.t. page faults is io_lock -> pagefault -> mmap_sem -> ilock. The directory readdir code now triggers ilock -> page fault -> mmap_sem. While we cannot deadlock at this point, it inverts all the locking patterns that lockdep normally sees on XFS inodes, and so triggers lockdep. We worked around this with commit 93a8614 ("xfs: fix directory inode iolock lockdep false positive"), but that then just moved the lockdep warning to deeper in the page fault path and triggered on security inode locks. Fixing the shmem issue there just moved the lockdep reports somewhere else, and now we are getting false positives from filesystem freezing annotations getting confused. Further, if we enter memory reclaim in a readdir path, we now get lockdep warning about potential deadlocks because the ilock is held when we enter reclaim. This, again, is different to a regular file in that we never allow memory reclaim to run while holding the ilock for regular files. Hence lockdep now throws ilock->kmalloc->reclaim->ilock warnings. Basically, the problem is that the ilock is being used to protect the directory data and the inode metadata, whereas for a regular file the iolock protects the data and the ilock protects the metadata. From the VFS perspective, the i_mutex serialises all accesses to the directory data, and so not holding the ilock for readdir doesn't matter. The issue is that CXFS doesn't access directory data via the VFS, so it has no "data serialisaton" mechanism. Hence we need to hold the IOLOCK in the correct places to provide this low level directory data access serialisation. The ilock can then be used just when the extent list needs to be read, just like we do for regular files. The directory modification code can take the iolock exclusive when the ilock is also taken, and this then ensures that readdir is correct excluded while modifications are in progress. Signed-off-by: Dave Chinner <dchinner@redhat.com> Reviewed-by: Brian Foster <bfoster@redhat.com> Signed-off-by: Dave Chinner <david@fromorbit.com>
2015-08-19 00:33:00 +00:00
lock_mode = xfs_ilock_data_map_shared(dp);
if (dp->i_df.if_format == XFS_DINODE_FMT_LOCAL) {
rval = xfs_dir2_sf_lookup(args);
goto out_check_rval;
}
rval = xfs_dir2_isblock(args, &v);
if (rval)
goto out_free;
if (v) {
rval = xfs_dir2_block_lookup(args);
goto out_check_rval;
}
rval = xfs_dir2_isleaf(args, &v);
if (rval)
goto out_free;
if (v)
rval = xfs_dir2_leaf_lookup(args);
else
rval = xfs_dir2_node_lookup(args);
out_check_rval:
if (rval == -EEXIST)
rval = 0;
if (!rval) {
*inum = args->inumber;
if (ci_name) {
ci_name->name = args->value;
ci_name->len = args->valuelen;
}
}
out_free:
xfs: stop holding ILOCK over filldir callbacks The recent change to the readdir locking made in 40194ec ("xfs: reinstate the ilock in xfs_readdir") for CXFS directory sanity was probably the wrong thing to do. Deep in the readdir code we can take page faults in the filldir callback, and so taking a page fault while holding an inode ilock creates a new set of locking issues that lockdep warns all over the place about. The locking order for regular inodes w.r.t. page faults is io_lock -> pagefault -> mmap_sem -> ilock. The directory readdir code now triggers ilock -> page fault -> mmap_sem. While we cannot deadlock at this point, it inverts all the locking patterns that lockdep normally sees on XFS inodes, and so triggers lockdep. We worked around this with commit 93a8614 ("xfs: fix directory inode iolock lockdep false positive"), but that then just moved the lockdep warning to deeper in the page fault path and triggered on security inode locks. Fixing the shmem issue there just moved the lockdep reports somewhere else, and now we are getting false positives from filesystem freezing annotations getting confused. Further, if we enter memory reclaim in a readdir path, we now get lockdep warning about potential deadlocks because the ilock is held when we enter reclaim. This, again, is different to a regular file in that we never allow memory reclaim to run while holding the ilock for regular files. Hence lockdep now throws ilock->kmalloc->reclaim->ilock warnings. Basically, the problem is that the ilock is being used to protect the directory data and the inode metadata, whereas for a regular file the iolock protects the data and the ilock protects the metadata. From the VFS perspective, the i_mutex serialises all accesses to the directory data, and so not holding the ilock for readdir doesn't matter. The issue is that CXFS doesn't access directory data via the VFS, so it has no "data serialisaton" mechanism. Hence we need to hold the IOLOCK in the correct places to provide this low level directory data access serialisation. The ilock can then be used just when the extent list needs to be read, just like we do for regular files. The directory modification code can take the iolock exclusive when the ilock is also taken, and this then ensures that readdir is correct excluded while modifications are in progress. Signed-off-by: Dave Chinner <dchinner@redhat.com> Reviewed-by: Brian Foster <bfoster@redhat.com> Signed-off-by: Dave Chinner <david@fromorbit.com>
2015-08-19 00:33:00 +00:00
xfs_iunlock(dp, lock_mode);
kmem_free(args);
return rval;
}
/*
* Remove an entry from a directory.
*/
int
xfs_dir_removename(
struct xfs_trans *tp,
struct xfs_inode *dp,
struct xfs_name *name,
xfs_ino_t ino,
xfs_extlen_t total) /* bmap's total block count */
{
struct xfs_da_args *args;
int rval;
bool v;
ASSERT(S_ISDIR(VFS_I(dp)->i_mode));
XFS_STATS_INC(dp->i_mount, xs_dir_remove);
args = kmem_zalloc(sizeof(*args), KM_NOFS);
if (!args)
return -ENOMEM;
args->geo = dp->i_mount->m_dir_geo;
args->name = name->name;
args->namelen = name->len;
args->filetype = name->type;
args->hashval = xfs_dir2_hashname(dp->i_mount, name);
args->inumber = ino;
args->dp = dp;
args->total = total;
args->whichfork = XFS_DATA_FORK;
args->trans = tp;
if (dp->i_df.if_format == XFS_DINODE_FMT_LOCAL) {
rval = xfs_dir2_sf_removename(args);
goto out_free;
}
rval = xfs_dir2_isblock(args, &v);
if (rval)
goto out_free;
if (v) {
rval = xfs_dir2_block_removename(args);
goto out_free;
}
rval = xfs_dir2_isleaf(args, &v);
if (rval)
goto out_free;
if (v)
rval = xfs_dir2_leaf_removename(args);
else
rval = xfs_dir2_node_removename(args);
out_free:
kmem_free(args);
return rval;
}
/*
* Replace the inode number of a directory entry.
*/
int
xfs_dir_replace(
struct xfs_trans *tp,
struct xfs_inode *dp,
const struct xfs_name *name, /* name of entry to replace */
xfs_ino_t inum, /* new inode number */
xfs_extlen_t total) /* bmap's total block count */
{
struct xfs_da_args *args;
int rval;
bool v;
ASSERT(S_ISDIR(VFS_I(dp)->i_mode));
rval = xfs_dir_ino_validate(tp->t_mountp, inum);
if (rval)
return rval;
args = kmem_zalloc(sizeof(*args), KM_NOFS);
if (!args)
return -ENOMEM;
args->geo = dp->i_mount->m_dir_geo;
args->name = name->name;
args->namelen = name->len;
args->filetype = name->type;
args->hashval = xfs_dir2_hashname(dp->i_mount, name);
args->inumber = inum;
args->dp = dp;
args->total = total;
args->whichfork = XFS_DATA_FORK;
args->trans = tp;
if (dp->i_df.if_format == XFS_DINODE_FMT_LOCAL) {
rval = xfs_dir2_sf_replace(args);
goto out_free;
}
rval = xfs_dir2_isblock(args, &v);
if (rval)
goto out_free;
if (v) {
rval = xfs_dir2_block_replace(args);
goto out_free;
}
rval = xfs_dir2_isleaf(args, &v);
if (rval)
goto out_free;
if (v)
rval = xfs_dir2_leaf_replace(args);
else
rval = xfs_dir2_node_replace(args);
out_free:
kmem_free(args);
return rval;
}
/*
* See if this entry can be added to the directory without allocating space.
*/
int
xfs_dir_canenter(
xfs_trans_t *tp,
xfs_inode_t *dp,
struct xfs_name *name) /* name of entry to add */
{
return xfs_dir_createname(tp, dp, name, 0, 0);
}
/*
* Utility routines.
*/
/*
* Add a block to the directory.
*
* This routine is for data and free blocks, not leaf/node blocks which are
* handled by xfs_da_grow_inode.
*/
int
xfs_dir2_grow_inode(
struct xfs_da_args *args,
int space, /* v2 dir's space XFS_DIR2_xxx_SPACE */
xfs_dir2_db_t *dbp) /* out: block number added */
{
struct xfs_inode *dp = args->dp;
struct xfs_mount *mp = dp->i_mount;
xfs_fileoff_t bno; /* directory offset of new block */
int count; /* count of filesystem blocks */
int error;
xfs: event tracing support Convert the old xfs tracing support that could only be used with the out of tree kdb and xfsidbg patches to use the generic event tracer. To use it make sure CONFIG_EVENT_TRACING is enabled and then enable all xfs trace channels by: echo 1 > /sys/kernel/debug/tracing/events/xfs/enable or alternatively enable single events by just doing the same in one event subdirectory, e.g. echo 1 > /sys/kernel/debug/tracing/events/xfs/xfs_ihold/enable or set more complex filters, etc. In Documentation/trace/events.txt all this is desctribed in more detail. To reads the events do a cat /sys/kernel/debug/tracing/trace Compared to the last posting this patch converts the tracing mostly to the one tracepoint per callsite model that other users of the new tracing facility also employ. This allows a very fine-grained control of the tracing, a cleaner output of the traces and also enables the perf tool to use each tracepoint as a virtual performance counter, allowing us to e.g. count how often certain workloads git various spots in XFS. Take a look at http://lwn.net/Articles/346470/ for some examples. Also the btree tracing isn't included at all yet, as it will require additional core tracing features not in mainline yet, I plan to deliver it later. And the really nice thing about this patch is that it actually removes many lines of code while adding this nice functionality: fs/xfs/Makefile | 8 fs/xfs/linux-2.6/xfs_acl.c | 1 fs/xfs/linux-2.6/xfs_aops.c | 52 - fs/xfs/linux-2.6/xfs_aops.h | 2 fs/xfs/linux-2.6/xfs_buf.c | 117 +-- fs/xfs/linux-2.6/xfs_buf.h | 33 fs/xfs/linux-2.6/xfs_fs_subr.c | 3 fs/xfs/linux-2.6/xfs_ioctl.c | 1 fs/xfs/linux-2.6/xfs_ioctl32.c | 1 fs/xfs/linux-2.6/xfs_iops.c | 1 fs/xfs/linux-2.6/xfs_linux.h | 1 fs/xfs/linux-2.6/xfs_lrw.c | 87 -- fs/xfs/linux-2.6/xfs_lrw.h | 45 - fs/xfs/linux-2.6/xfs_super.c | 104 --- fs/xfs/linux-2.6/xfs_super.h | 7 fs/xfs/linux-2.6/xfs_sync.c | 1 fs/xfs/linux-2.6/xfs_trace.c | 75 ++ fs/xfs/linux-2.6/xfs_trace.h | 1369 +++++++++++++++++++++++++++++++++++++++++ fs/xfs/linux-2.6/xfs_vnode.h | 4 fs/xfs/quota/xfs_dquot.c | 110 --- fs/xfs/quota/xfs_dquot.h | 21 fs/xfs/quota/xfs_qm.c | 40 - fs/xfs/quota/xfs_qm_syscalls.c | 4 fs/xfs/support/ktrace.c | 323 --------- fs/xfs/support/ktrace.h | 85 -- fs/xfs/xfs.h | 16 fs/xfs/xfs_ag.h | 14 fs/xfs/xfs_alloc.c | 230 +----- fs/xfs/xfs_alloc.h | 27 fs/xfs/xfs_alloc_btree.c | 1 fs/xfs/xfs_attr.c | 107 --- fs/xfs/xfs_attr.h | 10 fs/xfs/xfs_attr_leaf.c | 14 fs/xfs/xfs_attr_sf.h | 40 - fs/xfs/xfs_bmap.c | 507 +++------------ fs/xfs/xfs_bmap.h | 49 - fs/xfs/xfs_bmap_btree.c | 6 fs/xfs/xfs_btree.c | 5 fs/xfs/xfs_btree_trace.h | 17 fs/xfs/xfs_buf_item.c | 87 -- fs/xfs/xfs_buf_item.h | 20 fs/xfs/xfs_da_btree.c | 3 fs/xfs/xfs_da_btree.h | 7 fs/xfs/xfs_dfrag.c | 2 fs/xfs/xfs_dir2.c | 8 fs/xfs/xfs_dir2_block.c | 20 fs/xfs/xfs_dir2_leaf.c | 21 fs/xfs/xfs_dir2_node.c | 27 fs/xfs/xfs_dir2_sf.c | 26 fs/xfs/xfs_dir2_trace.c | 216 ------ fs/xfs/xfs_dir2_trace.h | 72 -- fs/xfs/xfs_filestream.c | 8 fs/xfs/xfs_fsops.c | 2 fs/xfs/xfs_iget.c | 111 --- fs/xfs/xfs_inode.c | 67 -- fs/xfs/xfs_inode.h | 76 -- fs/xfs/xfs_inode_item.c | 5 fs/xfs/xfs_iomap.c | 85 -- fs/xfs/xfs_iomap.h | 8 fs/xfs/xfs_log.c | 181 +---- fs/xfs/xfs_log_priv.h | 20 fs/xfs/xfs_log_recover.c | 1 fs/xfs/xfs_mount.c | 2 fs/xfs/xfs_quota.h | 8 fs/xfs/xfs_rename.c | 1 fs/xfs/xfs_rtalloc.c | 1 fs/xfs/xfs_rw.c | 3 fs/xfs/xfs_trans.h | 47 + fs/xfs/xfs_trans_buf.c | 62 - fs/xfs/xfs_vnodeops.c | 8 70 files changed, 2151 insertions(+), 2592 deletions(-) Signed-off-by: Christoph Hellwig <hch@lst.de> Signed-off-by: Alex Elder <aelder@sgi.com>
2009-12-14 23:14:59 +00:00
trace_xfs_dir2_grow_inode(args, space);
/*
* Set lowest possible block in the space requested.
*/
bno = XFS_B_TO_FSBT(mp, space * XFS_DIR2_SPACE_SIZE);
count = args->geo->fsbcount;
error = xfs_da_grow_inode_int(args, &bno, count);
if (error)
return error;
2008-10-30 06:38:12 +00:00
*dbp = xfs_dir2_da_to_db(args->geo, (xfs_dablk_t)bno);
2008-10-30 06:38:12 +00:00
/*
* Update file's size if this is the data space and it grew.
*/
if (space == XFS_DIR2_DATA_SPACE) {
xfs_fsize_t size; /* directory file (data) size */
size = XFS_FSB_TO_B(mp, bno + count);
if (size > dp->i_disk_size) {
dp->i_disk_size = size;
xfs_trans_log_inode(args->trans, dp, XFS_ILOG_CORE);
}
}
return 0;
}
/*
* See if the directory is a single-block form directory.
*/
int
xfs_dir2_isblock(
struct xfs_da_args *args,
bool *isblock)
{
struct xfs_mount *mp = args->dp->i_mount;
xfs_fileoff_t eof;
int error;
error = xfs_bmap_last_offset(args->dp, &eof, XFS_DATA_FORK);
if (error)
return error;
*isblock = false;
if (XFS_FSB_TO_B(mp, eof) != args->geo->blksize)
return 0;
*isblock = true;
if (XFS_IS_CORRUPT(mp, args->dp->i_disk_size != args->geo->blksize))
return -EFSCORRUPTED;
return 0;
}
/*
* See if the directory is a single-leaf form directory.
*/
int
xfs_dir2_isleaf(
struct xfs_da_args *args,
bool *isleaf)
{
xfs_fileoff_t eof;
int error;
error = xfs_bmap_last_offset(args->dp, &eof, XFS_DATA_FORK);
if (error)
return error;
*isleaf = false;
if (eof != args->geo->leafblk + args->geo->fsbcount)
return 0;
*isleaf = true;
return 0;
}
/*
* Remove the given block from the directory.
* This routine is used for data and free blocks, leaf/node are done
* by xfs_da_shrink_inode.
*/
int
xfs_dir2_shrink_inode(
struct xfs_da_args *args,
xfs_dir2_db_t db,
struct xfs_buf *bp)
{
xfs_fileoff_t bno; /* directory file offset */
xfs_dablk_t da; /* directory file offset */
int done; /* bunmap is finished */
struct xfs_inode *dp;
int error;
struct xfs_mount *mp;
struct xfs_trans *tp;
xfs: event tracing support Convert the old xfs tracing support that could only be used with the out of tree kdb and xfsidbg patches to use the generic event tracer. To use it make sure CONFIG_EVENT_TRACING is enabled and then enable all xfs trace channels by: echo 1 > /sys/kernel/debug/tracing/events/xfs/enable or alternatively enable single events by just doing the same in one event subdirectory, e.g. echo 1 > /sys/kernel/debug/tracing/events/xfs/xfs_ihold/enable or set more complex filters, etc. In Documentation/trace/events.txt all this is desctribed in more detail. To reads the events do a cat /sys/kernel/debug/tracing/trace Compared to the last posting this patch converts the tracing mostly to the one tracepoint per callsite model that other users of the new tracing facility also employ. This allows a very fine-grained control of the tracing, a cleaner output of the traces and also enables the perf tool to use each tracepoint as a virtual performance counter, allowing us to e.g. count how often certain workloads git various spots in XFS. Take a look at http://lwn.net/Articles/346470/ for some examples. Also the btree tracing isn't included at all yet, as it will require additional core tracing features not in mainline yet, I plan to deliver it later. And the really nice thing about this patch is that it actually removes many lines of code while adding this nice functionality: fs/xfs/Makefile | 8 fs/xfs/linux-2.6/xfs_acl.c | 1 fs/xfs/linux-2.6/xfs_aops.c | 52 - fs/xfs/linux-2.6/xfs_aops.h | 2 fs/xfs/linux-2.6/xfs_buf.c | 117 +-- fs/xfs/linux-2.6/xfs_buf.h | 33 fs/xfs/linux-2.6/xfs_fs_subr.c | 3 fs/xfs/linux-2.6/xfs_ioctl.c | 1 fs/xfs/linux-2.6/xfs_ioctl32.c | 1 fs/xfs/linux-2.6/xfs_iops.c | 1 fs/xfs/linux-2.6/xfs_linux.h | 1 fs/xfs/linux-2.6/xfs_lrw.c | 87 -- fs/xfs/linux-2.6/xfs_lrw.h | 45 - fs/xfs/linux-2.6/xfs_super.c | 104 --- fs/xfs/linux-2.6/xfs_super.h | 7 fs/xfs/linux-2.6/xfs_sync.c | 1 fs/xfs/linux-2.6/xfs_trace.c | 75 ++ fs/xfs/linux-2.6/xfs_trace.h | 1369 +++++++++++++++++++++++++++++++++++++++++ fs/xfs/linux-2.6/xfs_vnode.h | 4 fs/xfs/quota/xfs_dquot.c | 110 --- fs/xfs/quota/xfs_dquot.h | 21 fs/xfs/quota/xfs_qm.c | 40 - fs/xfs/quota/xfs_qm_syscalls.c | 4 fs/xfs/support/ktrace.c | 323 --------- fs/xfs/support/ktrace.h | 85 -- fs/xfs/xfs.h | 16 fs/xfs/xfs_ag.h | 14 fs/xfs/xfs_alloc.c | 230 +----- fs/xfs/xfs_alloc.h | 27 fs/xfs/xfs_alloc_btree.c | 1 fs/xfs/xfs_attr.c | 107 --- fs/xfs/xfs_attr.h | 10 fs/xfs/xfs_attr_leaf.c | 14 fs/xfs/xfs_attr_sf.h | 40 - fs/xfs/xfs_bmap.c | 507 +++------------ fs/xfs/xfs_bmap.h | 49 - fs/xfs/xfs_bmap_btree.c | 6 fs/xfs/xfs_btree.c | 5 fs/xfs/xfs_btree_trace.h | 17 fs/xfs/xfs_buf_item.c | 87 -- fs/xfs/xfs_buf_item.h | 20 fs/xfs/xfs_da_btree.c | 3 fs/xfs/xfs_da_btree.h | 7 fs/xfs/xfs_dfrag.c | 2 fs/xfs/xfs_dir2.c | 8 fs/xfs/xfs_dir2_block.c | 20 fs/xfs/xfs_dir2_leaf.c | 21 fs/xfs/xfs_dir2_node.c | 27 fs/xfs/xfs_dir2_sf.c | 26 fs/xfs/xfs_dir2_trace.c | 216 ------ fs/xfs/xfs_dir2_trace.h | 72 -- fs/xfs/xfs_filestream.c | 8 fs/xfs/xfs_fsops.c | 2 fs/xfs/xfs_iget.c | 111 --- fs/xfs/xfs_inode.c | 67 -- fs/xfs/xfs_inode.h | 76 -- fs/xfs/xfs_inode_item.c | 5 fs/xfs/xfs_iomap.c | 85 -- fs/xfs/xfs_iomap.h | 8 fs/xfs/xfs_log.c | 181 +---- fs/xfs/xfs_log_priv.h | 20 fs/xfs/xfs_log_recover.c | 1 fs/xfs/xfs_mount.c | 2 fs/xfs/xfs_quota.h | 8 fs/xfs/xfs_rename.c | 1 fs/xfs/xfs_rtalloc.c | 1 fs/xfs/xfs_rw.c | 3 fs/xfs/xfs_trans.h | 47 + fs/xfs/xfs_trans_buf.c | 62 - fs/xfs/xfs_vnodeops.c | 8 70 files changed, 2151 insertions(+), 2592 deletions(-) Signed-off-by: Christoph Hellwig <hch@lst.de> Signed-off-by: Alex Elder <aelder@sgi.com>
2009-12-14 23:14:59 +00:00
trace_xfs_dir2_shrink_inode(args, db);
dp = args->dp;
mp = dp->i_mount;
tp = args->trans;
da = xfs_dir2_db_to_da(args->geo, db);
/* Unmap the fsblock(s). */
error = xfs_bunmapi(tp, dp, da, args->geo->fsbcount, 0, 0, &done);
if (error) {
/*
* ENOSPC actually can happen if we're in a removename with no
* space reservation, and the resulting block removal would
* cause a bmap btree split or conversion from extents to btree.
* This can only happen for un-fragmented directory blocks,
* since you need to be punching out the middle of an extent.
* In this case we need to leave the block in the file, and not
* binval it. So the block has to be in a consistent empty
* state and appropriately logged. We don't free up the buffer,
* the caller can tell it hasn't happened since it got an error
* back.
*/
return error;
}
ASSERT(done);
/*
* Invalidate the buffer from the transaction.
*/
xfs_trans_binval(tp, bp);
/*
* If it's not a data block, we're done.
*/
if (db >= xfs_dir2_byte_to_db(args->geo, XFS_DIR2_LEAF_OFFSET))
return 0;
/*
* If the block isn't the last one in the directory, we're done.
*/
if (dp->i_disk_size > xfs_dir2_db_off_to_byte(args->geo, db + 1, 0))
return 0;
bno = da;
if ((error = xfs_bmap_last_before(tp, dp, &bno, XFS_DATA_FORK))) {
/*
* This can't really happen unless there's kernel corruption.
*/
return error;
}
if (db == args->geo->datablk)
ASSERT(bno == 0);
else
ASSERT(bno > 0);
/*
* Set the size to the new last block.
*/
dp->i_disk_size = XFS_FSB_TO_B(mp, bno);
xfs_trans_log_inode(tp, dp, XFS_ILOG_CORE);
return 0;
}
/* Returns true if the directory entry name is valid. */
bool
xfs_dir2_namecheck(
const void *name,
size_t length)
{
/*
* MAXNAMELEN includes the trailing null, but (name/length) leave it
* out, so use >= for the length check.
*/
if (length >= MAXNAMELEN)
return false;
/* There shouldn't be any slashes or nulls here */
return !memchr(name, '/', length) && !memchr(name, 0, length);
}
xfs_dahash_t
xfs_dir2_hashname(
struct xfs_mount *mp,
const struct xfs_name *name)
{
if (unlikely(xfs_has_asciici(mp)))
return xfs_ascii_ci_hashname(name);
return xfs_da_hashname(name->name, name->len);
}
enum xfs_dacmp
xfs_dir2_compname(
struct xfs_da_args *args,
const unsigned char *name,
int len)
{
if (unlikely(xfs_has_asciici(args->dp->i_mount)))
return xfs_ascii_ci_compname(args, name, len);
return xfs_da_compname(args, name, len);
}