linux_dsm_epyc7002/fs/cifs/cifsfs.c

1149 lines
30 KiB
C
Raw Normal View History

/*
* fs/cifs/cifsfs.c
*
* Copyright (C) International Business Machines Corp., 2002,2008
* Author(s): Steve French (sfrench@us.ibm.com)
*
* Common Internet FileSystem (CIFS) client
*
* This library is free software; you can redistribute it and/or modify
* it under the terms of the GNU Lesser General Public License as published
* by the Free Software Foundation; either version 2.1 of the License, or
* (at your option) any later version.
*
* This library is distributed in the hope that it will be useful,
* but WITHOUT ANY WARRANTY; without even the implied warranty of
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See
* the GNU Lesser General Public License for more details.
*
* You should have received a copy of the GNU Lesser General Public License
* along with this library; if not, write to the Free Software
* Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
*/
/* Note that BB means BUGBUG (ie something to fix eventually) */
#include <linux/module.h>
#include <linux/fs.h>
#include <linux/mount.h>
#include <linux/slab.h>
#include <linux/init.h>
#include <linux/list.h>
#include <linux/seq_file.h>
#include <linux/vfs.h>
#include <linux/mempool.h>
#include <linux/delay.h>
#include <linux/kthread.h>
#include <linux/freezer.h>
#include "cifsfs.h"
#include "cifspdu.h"
#define DECLARE_GLOBALS_HERE
#include "cifsglob.h"
#include "cifsproto.h"
#include "cifs_debug.h"
#include "cifs_fs_sb.h"
#include <linux/mm.h>
#include <linux/key-type.h>
#include "dns_resolve.h"
#include "cifs_spnego.h"
#define CIFS_MAGIC_NUMBER 0xFF534D42 /* the first four bytes of SMB PDUs */
#ifdef CONFIG_CIFS_QUOTA
static struct quotactl_ops cifs_quotactl_ops;
#endif /* QUOTA */
int cifsFYI = 0;
int cifsERROR = 1;
int traceSMB = 0;
unsigned int oplockEnabled = 1;
unsigned int experimEnabled = 0;
unsigned int linuxExtEnabled = 1;
unsigned int lookupCacheEnabled = 1;
unsigned int multiuser_mount = 0;
unsigned int extended_security = CIFSSEC_DEF;
/* unsigned int ntlmv2_support = 0; */
unsigned int sign_CIFS_PDUs = 1;
extern struct task_struct *oplockThread; /* remove sparse warning */
struct task_struct *oplockThread = NULL;
/* extern struct task_struct * dnotifyThread; remove sparse warning */
static struct task_struct *dnotifyThread = NULL;
static const struct super_operations cifs_super_ops;
unsigned int CIFSMaxBufSize = CIFS_MAX_MSGSIZE;
module_param(CIFSMaxBufSize, int, 0);
MODULE_PARM_DESC(CIFSMaxBufSize, "Network buffer size (not including header). "
"Default: 16384 Range: 8192 to 130048");
unsigned int cifs_min_rcv = CIFS_MIN_RCV_POOL;
module_param(cifs_min_rcv, int, 0);
MODULE_PARM_DESC(cifs_min_rcv, "Network buffers in pool. Default: 4 Range: "
"1 to 64");
unsigned int cifs_min_small = 30;
module_param(cifs_min_small, int, 0);
MODULE_PARM_DESC(cifs_min_small, "Small network buffers in pool. Default: 30 "
"Range: 2 to 256");
unsigned int cifs_max_pending = CIFS_MAX_REQ;
module_param(cifs_max_pending, int, 0);
MODULE_PARM_DESC(cifs_max_pending, "Simultaneous requests to server. "
"Default: 50 Range: 2 to 256");
extern mempool_t *cifs_sm_req_poolp;
extern mempool_t *cifs_req_poolp;
extern mempool_t *cifs_mid_poolp;
extern struct kmem_cache *cifs_oplock_cachep;
static int
cifs_read_super(struct super_block *sb, void *data,
const char *devname, int silent)
{
struct inode *inode;
struct cifs_sb_info *cifs_sb;
#ifdef CONFIG_CIFS_DFS_UPCALL
int len;
#endif
int rc = 0;
/* BB should we make this contingent on mount parm? */
sb->s_flags |= MS_NODIRATIME | MS_NOATIME;
sb->s_fs_info = kzalloc(sizeof(struct cifs_sb_info), GFP_KERNEL);
cifs_sb = CIFS_SB(sb);
if (cifs_sb == NULL)
return -ENOMEM;
#ifdef CONFIG_CIFS_DFS_UPCALL
/* copy mount params to sb for use in submounts */
/* BB: should we move this after the mount so we
* do not have to do the copy on failed mounts?
* BB: May be it is better to do simple copy before
* complex operation (mount), and in case of fail
* just exit instead of doing mount and attempting
* undo it if this copy fails?*/
len = strlen(data);
cifs_sb->mountdata = kzalloc(len + 1, GFP_KERNEL);
if (cifs_sb->mountdata == NULL) {
kfree(sb->s_fs_info);
sb->s_fs_info = NULL;
return -ENOMEM;
}
strncpy(cifs_sb->mountdata, data, len + 1);
cifs_sb->mountdata[len] = '\0';
#endif
rc = cifs_mount(sb, cifs_sb, data, devname);
if (rc) {
if (!silent)
cERROR(1,
("cifs_mount failed w/return code = %d", rc));
goto out_mount_failed;
}
sb->s_magic = CIFS_MAGIC_NUMBER;
sb->s_op = &cifs_super_ops;
/* if (cifs_sb->tcon->ses->server->maxBuf > MAX_CIFS_HDR_SIZE + 512)
sb->s_blocksize =
cifs_sb->tcon->ses->server->maxBuf - MAX_CIFS_HDR_SIZE; */
#ifdef CONFIG_CIFS_QUOTA
sb->s_qcop = &cifs_quotactl_ops;
#endif
sb->s_blocksize = CIFS_MAX_MSGSIZE;
sb->s_blocksize_bits = 14; /* default 2**14 = CIFS_MAX_MSGSIZE */
inode = cifs_iget(sb, ROOT_I);
if (IS_ERR(inode)) {
rc = PTR_ERR(inode);
inode = NULL;
goto out_no_root;
}
sb->s_root = d_alloc_root(inode);
if (!sb->s_root) {
rc = -ENOMEM;
goto out_no_root;
}
#ifdef CONFIG_CIFS_EXPERIMENTAL
if (cifs_sb->mnt_cifs_flags & CIFS_MOUNT_SERVER_INUM) {
cFYI(1, ("export ops supported"));
sb->s_export_op = &cifs_export_ops;
}
#endif /* EXPERIMENTAL */
return 0;
out_no_root:
cERROR(1, ("cifs_read_super: get root inode failed"));
if (inode)
iput(inode);
out_mount_failed:
if (cifs_sb) {
#ifdef CONFIG_CIFS_DFS_UPCALL
if (cifs_sb->mountdata) {
kfree(cifs_sb->mountdata);
cifs_sb->mountdata = NULL;
}
#endif
if (cifs_sb->local_nls)
unload_nls(cifs_sb->local_nls);
kfree(cifs_sb);
}
return rc;
}
static void
cifs_put_super(struct super_block *sb)
{
int rc = 0;
struct cifs_sb_info *cifs_sb;
cFYI(1, ("In cifs_put_super"));
cifs_sb = CIFS_SB(sb);
if (cifs_sb == NULL) {
cFYI(1, ("Empty cifs superblock info passed to unmount"));
return;
}
rc = cifs_umount(sb, cifs_sb);
if (rc)
cERROR(1, ("cifs_umount failed with return code %d", rc));
#ifdef CONFIG_CIFS_DFS_UPCALL
if (cifs_sb->mountdata) {
kfree(cifs_sb->mountdata);
cifs_sb->mountdata = NULL;
}
#endif
unload_nls(cifs_sb->local_nls);
kfree(cifs_sb);
return;
}
static int
cifs_statfs(struct dentry *dentry, struct kstatfs *buf)
{
struct super_block *sb = dentry->d_sb;
struct cifs_sb_info *cifs_sb = CIFS_SB(sb);
struct cifsTconInfo *tcon = cifs_sb->tcon;
int rc = -EOPNOTSUPP;
int xid;
xid = GetXid();
buf->f_type = CIFS_MAGIC_NUMBER;
/*
* PATH_MAX may be too long - it would presumably be total path,
* but note that some servers (includinng Samba 3) have a shorter
* maximum path.
*
* Instead could get the real value via SMB_QUERY_FS_ATTRIBUTE_INFO.
*/
buf->f_namelen = PATH_MAX;
buf->f_files = 0; /* undefined */
buf->f_ffree = 0; /* unlimited */
/*
* We could add a second check for a QFS Unix capability bit
*/
if ((tcon->ses->capabilities & CAP_UNIX) &&
(CIFS_POSIX_EXTENSIONS & le64_to_cpu(tcon->fsUnixInfo.Capability)))
rc = CIFSSMBQFSPosixInfo(xid, tcon, buf);
/*
* Only need to call the old QFSInfo if failed on newer one,
* e.g. by OS/2.
**/
if (rc && (tcon->ses->capabilities & CAP_NT_SMBS))
rc = CIFSSMBQFSInfo(xid, tcon, buf);
/*
* Some old Windows servers also do not support level 103, retry with
* older level one if old server failed the previous call or we
* bypassed it because we detected that this was an older LANMAN sess
*/
if (rc)
rc = SMBOldQFSInfo(xid, tcon, buf);
FreeXid(xid);
return 0;
}
static int cifs_permission(struct inode *inode, int mask, struct nameidata *nd)
{
struct cifs_sb_info *cifs_sb;
cifs_sb = CIFS_SB(inode->i_sb);
if (cifs_sb->mnt_cifs_flags & CIFS_MOUNT_NO_PERM)
return 0;
else /* file mode might have been restricted at mount time
on the client (above and beyond ACL on servers) for
servers which do not support setting and viewing mode bits,
so allowing client to check permissions is useful */
return generic_permission(inode, mask, NULL);
}
static struct kmem_cache *cifs_inode_cachep;
static struct kmem_cache *cifs_req_cachep;
static struct kmem_cache *cifs_mid_cachep;
struct kmem_cache *cifs_oplock_cachep;
static struct kmem_cache *cifs_sm_req_cachep;
mempool_t *cifs_sm_req_poolp;
mempool_t *cifs_req_poolp;
mempool_t *cifs_mid_poolp;
static struct inode *
cifs_alloc_inode(struct super_block *sb)
{
struct cifsInodeInfo *cifs_inode;
cifs_inode = kmem_cache_alloc(cifs_inode_cachep, GFP_KERNEL);
if (!cifs_inode)
return NULL;
cifs_inode->cifsAttrs = 0x20; /* default */
atomic_set(&cifs_inode->inUse, 0);
cifs_inode->time = 0;
[CIFS] Fix potential data corruption when writing out cached dirty pages Fix RedHat bug 329431 The idea here is separate "conscious" from "unconscious" flushes. Conscious flushes are those due to a fsync() or close(). Unconscious ones are flushes that occur as a side effect of some other operation or due to memory pressure. Currently, when an error occurs during an unconscious flush (ENOSPC or EIO), we toss out the page and don't preserve that error to report to the user when a conscious flush occurs. If after the unconscious flush, there are no more dirty pages for the inode, the conscious flush will simply return success even though there were previous errors when writing out pages. This can lead to data corruption. The easiest way to reproduce this is to mount up a CIFS share that's very close to being full or where the user is very close to quota. mv a file to the share that's slightly larger than the quota allows. The writes will all succeed (since they go to pagecache). The mv will do a setattr to set the new file's attributes. This calls filemap_write_and_wait, which will return an error since all of the pages can't be written out. Then later, when the flush and release ops occur, there are no more dirty pages in pagecache for the file and those operations return 0. mv then assumes that the file was written out correctly and deletes the original. CIFS already has a write_behind_rc variable where it stores the results from earlier flushes, but that value is only reported in cifs_close. Since the VFS ignores the return value from the release operation, this isn't helpful. We should be reporting this error during the flush operation. This patch does the following: 1) changes cifs_fsync to use filemap_write_and_wait and cifs_flush and also sync to check its return code. If it returns successful, they then check the value of write_behind_rc to see if an earlier flush had reported any errors. If so, they return that error and clear write_behind_rc. 2) sets write_behind_rc in a few other places where pages are written out as a side effect of other operations and the code waits on them. 3) changes cifs_setattr to only call filemap_write_and_wait for ATTR_SIZE changes. 4) makes cifs_writepages accurately distinguish between EIO and ENOSPC errors when writing out pages. Some simple testing indicates that the patch works as expected and that it fixes the reproduceable known problem. Acked-by: Dave Kleikamp <shaggy@austin.rr.com> Signed-off-by: Jeff Layton <jlayton@redhat.com> Signed-off-by: Steve French <sfrench@us.ibm.com>
2007-11-21 06:19:03 +07:00
cifs_inode->write_behind_rc = 0;
/* Until the file is open and we have gotten oplock
info back from the server, can not assume caching of
file data or metadata */
cifs_inode->clientCanCacheRead = false;
cifs_inode->clientCanCacheAll = false;
cifs_inode->vfs_inode.i_blkbits = 14; /* 2**14 = CIFS_MAX_MSGSIZE */
/* Can not set i_flags here - they get immediately overwritten
to zero by the VFS */
/* cifs_inode->vfs_inode.i_flags = S_NOATIME | S_NOCMTIME;*/
INIT_LIST_HEAD(&cifs_inode->openFileList);
return &cifs_inode->vfs_inode;
}
static void
cifs_destroy_inode(struct inode *inode)
{
kmem_cache_free(cifs_inode_cachep, CIFS_I(inode));
}
/*
* cifs_show_options() is for displaying mount options in /proc/mounts.
* Not all settable options are displayed but most of the important
* ones are.
*/
static int
cifs_show_options(struct seq_file *s, struct vfsmount *m)
{
struct cifs_sb_info *cifs_sb;
cifs_sb = CIFS_SB(m->mnt_sb);
if (cifs_sb) {
if (cifs_sb->tcon) {
/* BB add prepath to mount options displayed */
seq_printf(s, ",unc=%s", cifs_sb->tcon->treeName);
if (cifs_sb->tcon->ses) {
if (cifs_sb->tcon->ses->userName)
seq_printf(s, ",username=%s",
cifs_sb->tcon->ses->userName);
if (cifs_sb->tcon->ses->domainName)
seq_printf(s, ",domain=%s",
cifs_sb->tcon->ses->domainName);
}
if ((cifs_sb->mnt_cifs_flags & CIFS_MOUNT_OVERR_UID) ||
!(cifs_sb->tcon->unix_ext))
seq_printf(s, ",uid=%d", cifs_sb->mnt_uid);
if ((cifs_sb->mnt_cifs_flags & CIFS_MOUNT_OVERR_GID) ||
!(cifs_sb->tcon->unix_ext))
seq_printf(s, ",gid=%d", cifs_sb->mnt_gid);
if (!cifs_sb->tcon->unix_ext) {
seq_printf(s, ",file_mode=0%o,dir_mode=0%o",
cifs_sb->mnt_file_mode,
cifs_sb->mnt_dir_mode);
}
if (cifs_sb->tcon->seal)
seq_printf(s, ",seal");
seq_printf(s, ",nocase");
}
if (cifs_sb->mnt_cifs_flags & CIFS_MOUNT_POSIX_PATHS)
seq_printf(s, ",posixpaths");
if (cifs_sb->mnt_cifs_flags & CIFS_MOUNT_SET_UID)
seq_printf(s, ",setuids");
if (cifs_sb->mnt_cifs_flags & CIFS_MOUNT_SERVER_INUM)
seq_printf(s, ",serverino");
if (cifs_sb->mnt_cifs_flags & CIFS_MOUNT_DIRECT_IO)
seq_printf(s, ",directio");
if (cifs_sb->mnt_cifs_flags & CIFS_MOUNT_NO_XATTR)
seq_printf(s, ",nouser_xattr");
if (cifs_sb->mnt_cifs_flags & CIFS_MOUNT_MAP_SPECIAL_CHR)
seq_printf(s, ",mapchars");
if (cifs_sb->mnt_cifs_flags & CIFS_MOUNT_UNX_EMUL)
seq_printf(s, ",sfu");
if (cifs_sb->mnt_cifs_flags & CIFS_MOUNT_NO_BRL)
seq_printf(s, ",nobrl");
if (cifs_sb->mnt_cifs_flags & CIFS_MOUNT_CIFS_ACL)
seq_printf(s, ",cifsacl");
if (cifs_sb->mnt_cifs_flags & CIFS_MOUNT_DYNPERM)
seq_printf(s, ",dynperm");
if (m->mnt_sb->s_flags & MS_POSIXACL)
seq_printf(s, ",acl");
seq_printf(s, ",rsize=%d", cifs_sb->rsize);
seq_printf(s, ",wsize=%d", cifs_sb->wsize);
}
return 0;
}
#ifdef CONFIG_CIFS_QUOTA
int cifs_xquota_set(struct super_block *sb, int quota_type, qid_t qid,
struct fs_disk_quota *pdquota)
{
int xid;
int rc = 0;
struct cifs_sb_info *cifs_sb = CIFS_SB(sb);
struct cifsTconInfo *pTcon;
if (cifs_sb)
pTcon = cifs_sb->tcon;
else
return -EIO;
xid = GetXid();
if (pTcon) {
cFYI(1, ("set type: 0x%x id: %d", quota_type, qid));
} else {
rc = -EIO;
}
FreeXid(xid);
return rc;
}
int cifs_xquota_get(struct super_block *sb, int quota_type, qid_t qid,
struct fs_disk_quota *pdquota)
{
int xid;
int rc = 0;
struct cifs_sb_info *cifs_sb = CIFS_SB(sb);
struct cifsTconInfo *pTcon;
if (cifs_sb)
pTcon = cifs_sb->tcon;
else
return -EIO;
xid = GetXid();
if (pTcon) {
cFYI(1, ("set type: 0x%x id: %d", quota_type, qid));
} else {
rc = -EIO;
}
FreeXid(xid);
return rc;
}
int cifs_xstate_set(struct super_block *sb, unsigned int flags, int operation)
{
int xid;
int rc = 0;
struct cifs_sb_info *cifs_sb = CIFS_SB(sb);
struct cifsTconInfo *pTcon;
if (cifs_sb)
pTcon = cifs_sb->tcon;
else
return -EIO;
xid = GetXid();
if (pTcon) {
cFYI(1, ("flags: 0x%x operation: 0x%x", flags, operation));
} else {
rc = -EIO;
}
FreeXid(xid);
return rc;
}
int cifs_xstate_get(struct super_block *sb, struct fs_quota_stat *qstats)
{
int xid;
int rc = 0;
struct cifs_sb_info *cifs_sb = CIFS_SB(sb);
struct cifsTconInfo *pTcon;
if (cifs_sb) {
pTcon = cifs_sb->tcon;
} else {
return -EIO;
}
xid = GetXid();
if (pTcon) {
cFYI(1, ("pqstats %p", qstats));
} else {
rc = -EIO;
}
FreeXid(xid);
return rc;
}
static struct quotactl_ops cifs_quotactl_ops = {
.set_xquota = cifs_xquota_set,
.get_xquota = cifs_xquota_get,
.set_xstate = cifs_xstate_set,
.get_xstate = cifs_xstate_get,
};
#endif
static void cifs_umount_begin(struct super_block *sb)
{
struct cifs_sb_info *cifs_sb = CIFS_SB(sb);
struct cifsTconInfo *tcon;
if (cifs_sb == NULL)
return;
tcon = cifs_sb->tcon;
if (tcon == NULL)
return;
down(&tcon->tconSem);
if (atomic_read(&tcon->useCount) == 1)
tcon->tidStatus = CifsExiting;
up(&tcon->tconSem);
/* cancel_brl_requests(tcon); */ /* BB mark all brl mids as exiting */
2005-11-10 06:21:09 +07:00
/* cancel_notify_requests(tcon); */
if (tcon->ses && tcon->ses->server) {
cFYI(1, ("wake up tasks now - umount begin not complete"));
wake_up_all(&tcon->ses->server->request_q);
wake_up_all(&tcon->ses->server->response_q);
msleep(1); /* yield */
/* we have to kick the requests once more */
wake_up_all(&tcon->ses->server->response_q);
msleep(1);
}
/* BB FIXME - finish add checks for tidStatus BB */
return;
}
#ifdef CONFIG_CIFS_STATS2
static int cifs_show_stats(struct seq_file *s, struct vfsmount *mnt)
{
/* BB FIXME */
return 0;
}
#endif
static int cifs_remount(struct super_block *sb, int *flags, char *data)
{
*flags |= MS_NODIRATIME;
return 0;
}
static const struct super_operations cifs_super_ops = {
.put_super = cifs_put_super,
.statfs = cifs_statfs,
.alloc_inode = cifs_alloc_inode,
.destroy_inode = cifs_destroy_inode,
/* .drop_inode = generic_delete_inode,
.delete_inode = cifs_delete_inode, */ /* Do not need above two
functions unless later we add lazy close of inodes or unless the
kernel forgets to call us with the same number of releases (closes)
as opens */
.show_options = cifs_show_options,
2005-11-10 06:21:09 +07:00
.umount_begin = cifs_umount_begin,
.remount_fs = cifs_remount,
#ifdef CONFIG_CIFS_STATS2
.show_stats = cifs_show_stats,
#endif
};
[PATCH] VFS: Permit filesystem to override root dentry on mount Extend the get_sb() filesystem operation to take an extra argument that permits the VFS to pass in the target vfsmount that defines the mountpoint. The filesystem is then required to manually set the superblock and root dentry pointers. For most filesystems, this should be done with simple_set_mnt() which will set the superblock pointer and then set the root dentry to the superblock's s_root (as per the old default behaviour). The get_sb() op now returns an integer as there's now no need to return the superblock pointer. This patch permits a superblock to be implicitly shared amongst several mount points, such as can be done with NFS to avoid potential inode aliasing. In such a case, simple_set_mnt() would not be called, and instead the mnt_root and mnt_sb would be set directly. The patch also makes the following changes: (*) the get_sb_*() convenience functions in the core kernel now take a vfsmount pointer argument and return an integer, so most filesystems have to change very little. (*) If one of the convenience function is not used, then get_sb() should normally call simple_set_mnt() to instantiate the vfsmount. This will always return 0, and so can be tail-called from get_sb(). (*) generic_shutdown_super() now calls shrink_dcache_sb() to clean up the dcache upon superblock destruction rather than shrink_dcache_anon(). This is required because the superblock may now have multiple trees that aren't actually bound to s_root, but that still need to be cleaned up. The currently called functions assume that the whole tree is rooted at s_root, and that anonymous dentries are not the roots of trees which results in dentries being left unculled. However, with the way NFS superblock sharing are currently set to be implemented, these assumptions are violated: the root of the filesystem is simply a dummy dentry and inode (the real inode for '/' may well be inaccessible), and all the vfsmounts are rooted on anonymous[*] dentries with child trees. [*] Anonymous until discovered from another tree. (*) The documentation has been adjusted, including the additional bit of changing ext2_* into foo_* in the documentation. [akpm@osdl.org: convert ipath_fs, do other stuff] Signed-off-by: David Howells <dhowells@redhat.com> Acked-by: Al Viro <viro@zeniv.linux.org.uk> Cc: Nathan Scott <nathans@sgi.com> Cc: Roland Dreier <rolandd@cisco.com> Signed-off-by: Andrew Morton <akpm@osdl.org> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2006-06-23 16:02:57 +07:00
static int
cifs_get_sb(struct file_system_type *fs_type,
[PATCH] VFS: Permit filesystem to override root dentry on mount Extend the get_sb() filesystem operation to take an extra argument that permits the VFS to pass in the target vfsmount that defines the mountpoint. The filesystem is then required to manually set the superblock and root dentry pointers. For most filesystems, this should be done with simple_set_mnt() which will set the superblock pointer and then set the root dentry to the superblock's s_root (as per the old default behaviour). The get_sb() op now returns an integer as there's now no need to return the superblock pointer. This patch permits a superblock to be implicitly shared amongst several mount points, such as can be done with NFS to avoid potential inode aliasing. In such a case, simple_set_mnt() would not be called, and instead the mnt_root and mnt_sb would be set directly. The patch also makes the following changes: (*) the get_sb_*() convenience functions in the core kernel now take a vfsmount pointer argument and return an integer, so most filesystems have to change very little. (*) If one of the convenience function is not used, then get_sb() should normally call simple_set_mnt() to instantiate the vfsmount. This will always return 0, and so can be tail-called from get_sb(). (*) generic_shutdown_super() now calls shrink_dcache_sb() to clean up the dcache upon superblock destruction rather than shrink_dcache_anon(). This is required because the superblock may now have multiple trees that aren't actually bound to s_root, but that still need to be cleaned up. The currently called functions assume that the whole tree is rooted at s_root, and that anonymous dentries are not the roots of trees which results in dentries being left unculled. However, with the way NFS superblock sharing are currently set to be implemented, these assumptions are violated: the root of the filesystem is simply a dummy dentry and inode (the real inode for '/' may well be inaccessible), and all the vfsmounts are rooted on anonymous[*] dentries with child trees. [*] Anonymous until discovered from another tree. (*) The documentation has been adjusted, including the additional bit of changing ext2_* into foo_* in the documentation. [akpm@osdl.org: convert ipath_fs, do other stuff] Signed-off-by: David Howells <dhowells@redhat.com> Acked-by: Al Viro <viro@zeniv.linux.org.uk> Cc: Nathan Scott <nathans@sgi.com> Cc: Roland Dreier <rolandd@cisco.com> Signed-off-by: Andrew Morton <akpm@osdl.org> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2006-06-23 16:02:57 +07:00
int flags, const char *dev_name, void *data, struct vfsmount *mnt)
{
int rc;
struct super_block *sb = sget(fs_type, NULL, set_anon_super, NULL);
cFYI(1, ("Devname: %s flags: %d ", dev_name, flags));
if (IS_ERR(sb))
[PATCH] VFS: Permit filesystem to override root dentry on mount Extend the get_sb() filesystem operation to take an extra argument that permits the VFS to pass in the target vfsmount that defines the mountpoint. The filesystem is then required to manually set the superblock and root dentry pointers. For most filesystems, this should be done with simple_set_mnt() which will set the superblock pointer and then set the root dentry to the superblock's s_root (as per the old default behaviour). The get_sb() op now returns an integer as there's now no need to return the superblock pointer. This patch permits a superblock to be implicitly shared amongst several mount points, such as can be done with NFS to avoid potential inode aliasing. In such a case, simple_set_mnt() would not be called, and instead the mnt_root and mnt_sb would be set directly. The patch also makes the following changes: (*) the get_sb_*() convenience functions in the core kernel now take a vfsmount pointer argument and return an integer, so most filesystems have to change very little. (*) If one of the convenience function is not used, then get_sb() should normally call simple_set_mnt() to instantiate the vfsmount. This will always return 0, and so can be tail-called from get_sb(). (*) generic_shutdown_super() now calls shrink_dcache_sb() to clean up the dcache upon superblock destruction rather than shrink_dcache_anon(). This is required because the superblock may now have multiple trees that aren't actually bound to s_root, but that still need to be cleaned up. The currently called functions assume that the whole tree is rooted at s_root, and that anonymous dentries are not the roots of trees which results in dentries being left unculled. However, with the way NFS superblock sharing are currently set to be implemented, these assumptions are violated: the root of the filesystem is simply a dummy dentry and inode (the real inode for '/' may well be inaccessible), and all the vfsmounts are rooted on anonymous[*] dentries with child trees. [*] Anonymous until discovered from another tree. (*) The documentation has been adjusted, including the additional bit of changing ext2_* into foo_* in the documentation. [akpm@osdl.org: convert ipath_fs, do other stuff] Signed-off-by: David Howells <dhowells@redhat.com> Acked-by: Al Viro <viro@zeniv.linux.org.uk> Cc: Nathan Scott <nathans@sgi.com> Cc: Roland Dreier <rolandd@cisco.com> Signed-off-by: Andrew Morton <akpm@osdl.org> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2006-06-23 16:02:57 +07:00
return PTR_ERR(sb);
sb->s_flags = flags;
rc = cifs_read_super(sb, data, dev_name, flags & MS_SILENT ? 1 : 0);
if (rc) {
up_write(&sb->s_umount);
deactivate_super(sb);
[PATCH] VFS: Permit filesystem to override root dentry on mount Extend the get_sb() filesystem operation to take an extra argument that permits the VFS to pass in the target vfsmount that defines the mountpoint. The filesystem is then required to manually set the superblock and root dentry pointers. For most filesystems, this should be done with simple_set_mnt() which will set the superblock pointer and then set the root dentry to the superblock's s_root (as per the old default behaviour). The get_sb() op now returns an integer as there's now no need to return the superblock pointer. This patch permits a superblock to be implicitly shared amongst several mount points, such as can be done with NFS to avoid potential inode aliasing. In such a case, simple_set_mnt() would not be called, and instead the mnt_root and mnt_sb would be set directly. The patch also makes the following changes: (*) the get_sb_*() convenience functions in the core kernel now take a vfsmount pointer argument and return an integer, so most filesystems have to change very little. (*) If one of the convenience function is not used, then get_sb() should normally call simple_set_mnt() to instantiate the vfsmount. This will always return 0, and so can be tail-called from get_sb(). (*) generic_shutdown_super() now calls shrink_dcache_sb() to clean up the dcache upon superblock destruction rather than shrink_dcache_anon(). This is required because the superblock may now have multiple trees that aren't actually bound to s_root, but that still need to be cleaned up. The currently called functions assume that the whole tree is rooted at s_root, and that anonymous dentries are not the roots of trees which results in dentries being left unculled. However, with the way NFS superblock sharing are currently set to be implemented, these assumptions are violated: the root of the filesystem is simply a dummy dentry and inode (the real inode for '/' may well be inaccessible), and all the vfsmounts are rooted on anonymous[*] dentries with child trees. [*] Anonymous until discovered from another tree. (*) The documentation has been adjusted, including the additional bit of changing ext2_* into foo_* in the documentation. [akpm@osdl.org: convert ipath_fs, do other stuff] Signed-off-by: David Howells <dhowells@redhat.com> Acked-by: Al Viro <viro@zeniv.linux.org.uk> Cc: Nathan Scott <nathans@sgi.com> Cc: Roland Dreier <rolandd@cisco.com> Signed-off-by: Andrew Morton <akpm@osdl.org> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2006-06-23 16:02:57 +07:00
return rc;
}
sb->s_flags |= MS_ACTIVE;
[PATCH] VFS: Permit filesystem to override root dentry on mount Extend the get_sb() filesystem operation to take an extra argument that permits the VFS to pass in the target vfsmount that defines the mountpoint. The filesystem is then required to manually set the superblock and root dentry pointers. For most filesystems, this should be done with simple_set_mnt() which will set the superblock pointer and then set the root dentry to the superblock's s_root (as per the old default behaviour). The get_sb() op now returns an integer as there's now no need to return the superblock pointer. This patch permits a superblock to be implicitly shared amongst several mount points, such as can be done with NFS to avoid potential inode aliasing. In such a case, simple_set_mnt() would not be called, and instead the mnt_root and mnt_sb would be set directly. The patch also makes the following changes: (*) the get_sb_*() convenience functions in the core kernel now take a vfsmount pointer argument and return an integer, so most filesystems have to change very little. (*) If one of the convenience function is not used, then get_sb() should normally call simple_set_mnt() to instantiate the vfsmount. This will always return 0, and so can be tail-called from get_sb(). (*) generic_shutdown_super() now calls shrink_dcache_sb() to clean up the dcache upon superblock destruction rather than shrink_dcache_anon(). This is required because the superblock may now have multiple trees that aren't actually bound to s_root, but that still need to be cleaned up. The currently called functions assume that the whole tree is rooted at s_root, and that anonymous dentries are not the roots of trees which results in dentries being left unculled. However, with the way NFS superblock sharing are currently set to be implemented, these assumptions are violated: the root of the filesystem is simply a dummy dentry and inode (the real inode for '/' may well be inaccessible), and all the vfsmounts are rooted on anonymous[*] dentries with child trees. [*] Anonymous until discovered from another tree. (*) The documentation has been adjusted, including the additional bit of changing ext2_* into foo_* in the documentation. [akpm@osdl.org: convert ipath_fs, do other stuff] Signed-off-by: David Howells <dhowells@redhat.com> Acked-by: Al Viro <viro@zeniv.linux.org.uk> Cc: Nathan Scott <nathans@sgi.com> Cc: Roland Dreier <rolandd@cisco.com> Signed-off-by: Andrew Morton <akpm@osdl.org> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2006-06-23 16:02:57 +07:00
return simple_set_mnt(mnt, sb);
}
static ssize_t cifs_file_aio_write(struct kiocb *iocb, const struct iovec *iov,
unsigned long nr_segs, loff_t pos)
{
struct inode *inode = iocb->ki_filp->f_path.dentry->d_inode;
ssize_t written;
written = generic_file_aio_write(iocb, iov, nr_segs, pos);
if (!CIFS_I(inode)->clientCanCacheAll)
filemap_fdatawrite(inode->i_mapping);
return written;
}
static loff_t cifs_llseek(struct file *file, loff_t offset, int origin)
{
/* origin == SEEK_END => we must revalidate the cached file length */
if (origin == SEEK_END) {
int retval;
/* some applications poll for the file length in this strange
way so we must seek to end on non-oplocked files by
setting the revalidate time to zero */
CIFS_I(file->f_path.dentry->d_inode)->time = 0;
retval = cifs_revalidate(file->f_path.dentry);
if (retval < 0)
return (loff_t)retval;
}
return remote_llseek(file, offset, origin);
}
struct file_system_type cifs_fs_type = {
.owner = THIS_MODULE,
.name = "cifs",
.get_sb = cifs_get_sb,
.kill_sb = kill_anon_super,
/* .fs_flags */
};
const struct inode_operations cifs_dir_inode_ops = {
.create = cifs_create,
.lookup = cifs_lookup,
.getattr = cifs_getattr,
.unlink = cifs_unlink,
.link = cifs_hardlink,
.mkdir = cifs_mkdir,
.rmdir = cifs_rmdir,
.rename = cifs_rename,
.permission = cifs_permission,
/* revalidate:cifs_revalidate, */
.setattr = cifs_setattr,
.symlink = cifs_symlink,
.mknod = cifs_mknod,
#ifdef CONFIG_CIFS_XATTR
.setxattr = cifs_setxattr,
.getxattr = cifs_getxattr,
.listxattr = cifs_listxattr,
.removexattr = cifs_removexattr,
#endif
};
const struct inode_operations cifs_file_inode_ops = {
/* revalidate:cifs_revalidate, */
.setattr = cifs_setattr,
.getattr = cifs_getattr, /* do we need this anymore? */
.rename = cifs_rename,
.permission = cifs_permission,
#ifdef CONFIG_CIFS_XATTR
.setxattr = cifs_setxattr,
.getxattr = cifs_getxattr,
.listxattr = cifs_listxattr,
.removexattr = cifs_removexattr,
#endif
};
const struct inode_operations cifs_symlink_inode_ops = {
.readlink = generic_readlink,
.follow_link = cifs_follow_link,
.put_link = cifs_put_link,
.permission = cifs_permission,
/* BB add the following two eventually */
/* revalidate: cifs_revalidate,
setattr: cifs_notify_change, *//* BB do we need notify change */
#ifdef CONFIG_CIFS_XATTR
.setxattr = cifs_setxattr,
.getxattr = cifs_getxattr,
.listxattr = cifs_listxattr,
.removexattr = cifs_removexattr,
#endif
};
const struct file_operations cifs_file_ops = {
.read = do_sync_read,
.write = do_sync_write,
.aio_read = generic_file_aio_read,
.aio_write = cifs_file_aio_write,
.open = cifs_open,
.release = cifs_close,
.lock = cifs_lock,
.fsync = cifs_fsync,
.flush = cifs_flush,
.mmap = cifs_file_mmap,
.splice_read = generic_file_splice_read,
.llseek = cifs_llseek,
#ifdef CONFIG_CIFS_POSIX
.unlocked_ioctl = cifs_ioctl,
#endif /* CONFIG_CIFS_POSIX */
#ifdef CONFIG_CIFS_EXPERIMENTAL
.dir_notify = cifs_dir_notify,
#endif /* CONFIG_CIFS_EXPERIMENTAL */
};
const struct file_operations cifs_file_direct_ops = {
/* no mmap, no aio, no readv -
BB reevaluate whether they can be done with directio, no cache */
.read = cifs_user_read,
.write = cifs_user_write,
.open = cifs_open,
.release = cifs_close,
.lock = cifs_lock,
.fsync = cifs_fsync,
.flush = cifs_flush,
.splice_read = generic_file_splice_read,
#ifdef CONFIG_CIFS_POSIX
.unlocked_ioctl = cifs_ioctl,
#endif /* CONFIG_CIFS_POSIX */
.llseek = cifs_llseek,
#ifdef CONFIG_CIFS_EXPERIMENTAL
.dir_notify = cifs_dir_notify,
#endif /* CONFIG_CIFS_EXPERIMENTAL */
};
const struct file_operations cifs_file_nobrl_ops = {
.read = do_sync_read,
.write = do_sync_write,
.aio_read = generic_file_aio_read,
.aio_write = cifs_file_aio_write,
.open = cifs_open,
.release = cifs_close,
.fsync = cifs_fsync,
.flush = cifs_flush,
.mmap = cifs_file_mmap,
.splice_read = generic_file_splice_read,
.llseek = cifs_llseek,
#ifdef CONFIG_CIFS_POSIX
.unlocked_ioctl = cifs_ioctl,
#endif /* CONFIG_CIFS_POSIX */
#ifdef CONFIG_CIFS_EXPERIMENTAL
.dir_notify = cifs_dir_notify,
#endif /* CONFIG_CIFS_EXPERIMENTAL */
};
const struct file_operations cifs_file_direct_nobrl_ops = {
/* no mmap, no aio, no readv -
BB reevaluate whether they can be done with directio, no cache */
.read = cifs_user_read,
.write = cifs_user_write,
.open = cifs_open,
.release = cifs_close,
.fsync = cifs_fsync,
.flush = cifs_flush,
.splice_read = generic_file_splice_read,
#ifdef CONFIG_CIFS_POSIX
.unlocked_ioctl = cifs_ioctl,
#endif /* CONFIG_CIFS_POSIX */
.llseek = cifs_llseek,
#ifdef CONFIG_CIFS_EXPERIMENTAL
.dir_notify = cifs_dir_notify,
#endif /* CONFIG_CIFS_EXPERIMENTAL */
};
const struct file_operations cifs_dir_ops = {
.readdir = cifs_readdir,
.release = cifs_closedir,
.read = generic_read_dir,
#ifdef CONFIG_CIFS_EXPERIMENTAL
.dir_notify = cifs_dir_notify,
#endif /* CONFIG_CIFS_EXPERIMENTAL */
.unlocked_ioctl = cifs_ioctl,
};
static void
cifs_init_once(struct kmem_cache *cachep, void *inode)
{
struct cifsInodeInfo *cifsi = inode;
inode_init_once(&cifsi->vfs_inode);
INIT_LIST_HEAD(&cifsi->lockList);
}
static int
cifs_init_inodecache(void)
{
cifs_inode_cachep = kmem_cache_create("cifs_inode_cache",
sizeof(struct cifsInodeInfo),
0, (SLAB_RECLAIM_ACCOUNT|
SLAB_MEM_SPREAD),
cifs_init_once);
if (cifs_inode_cachep == NULL)
return -ENOMEM;
return 0;
}
static void
cifs_destroy_inodecache(void)
{
kmem_cache_destroy(cifs_inode_cachep);
}
static int
cifs_init_request_bufs(void)
{
if (CIFSMaxBufSize < 8192) {
/* Buffer size can not be smaller than 2 * PATH_MAX since maximum
Unicode path name has to fit in any SMB/CIFS path based frames */
CIFSMaxBufSize = 8192;
} else if (CIFSMaxBufSize > 1024*127) {
CIFSMaxBufSize = 1024 * 127;
} else {
CIFSMaxBufSize &= 0x1FE00; /* Round size to even 512 byte mult*/
}
/* cERROR(1,("CIFSMaxBufSize %d 0x%x",CIFSMaxBufSize,CIFSMaxBufSize)); */
cifs_req_cachep = kmem_cache_create("cifs_request",
CIFSMaxBufSize +
MAX_CIFS_HDR_SIZE, 0,
SLAB_HWCACHE_ALIGN, NULL);
if (cifs_req_cachep == NULL)
return -ENOMEM;
if (cifs_min_rcv < 1)
cifs_min_rcv = 1;
else if (cifs_min_rcv > 64) {
cifs_min_rcv = 64;
cERROR(1, ("cifs_min_rcv set to maximum (64)"));
}
cifs_req_poolp = mempool_create_slab_pool(cifs_min_rcv,
cifs_req_cachep);
if (cifs_req_poolp == NULL) {
kmem_cache_destroy(cifs_req_cachep);
return -ENOMEM;
}
/* MAX_CIFS_SMALL_BUFFER_SIZE bytes is enough for most SMB responses and
almost all handle based requests (but not write response, nor is it
sufficient for path based requests). A smaller size would have
been more efficient (compacting multiple slab items on one 4k page)
for the case in which debug was on, but this larger size allows
more SMBs to use small buffer alloc and is still much more
efficient to alloc 1 per page off the slab compared to 17K (5page)
alloc of large cifs buffers even when page debugging is on */
cifs_sm_req_cachep = kmem_cache_create("cifs_small_rq",
MAX_CIFS_SMALL_BUFFER_SIZE, 0, SLAB_HWCACHE_ALIGN,
NULL);
if (cifs_sm_req_cachep == NULL) {
mempool_destroy(cifs_req_poolp);
kmem_cache_destroy(cifs_req_cachep);
return -ENOMEM;
}
if (cifs_min_small < 2)
cifs_min_small = 2;
else if (cifs_min_small > 256) {
cifs_min_small = 256;
cFYI(1, ("cifs_min_small set to maximum (256)"));
}
cifs_sm_req_poolp = mempool_create_slab_pool(cifs_min_small,
cifs_sm_req_cachep);
if (cifs_sm_req_poolp == NULL) {
mempool_destroy(cifs_req_poolp);
kmem_cache_destroy(cifs_req_cachep);
kmem_cache_destroy(cifs_sm_req_cachep);
return -ENOMEM;
}
return 0;
}
static void
cifs_destroy_request_bufs(void)
{
mempool_destroy(cifs_req_poolp);
kmem_cache_destroy(cifs_req_cachep);
mempool_destroy(cifs_sm_req_poolp);
kmem_cache_destroy(cifs_sm_req_cachep);
}
static int
cifs_init_mids(void)
{
cifs_mid_cachep = kmem_cache_create("cifs_mpx_ids",
sizeof(struct mid_q_entry), 0,
SLAB_HWCACHE_ALIGN, NULL);
if (cifs_mid_cachep == NULL)
return -ENOMEM;
/* 3 is a reasonable minimum number of simultaneous operations */
cifs_mid_poolp = mempool_create_slab_pool(3, cifs_mid_cachep);
if (cifs_mid_poolp == NULL) {
kmem_cache_destroy(cifs_mid_cachep);
return -ENOMEM;
}
cifs_oplock_cachep = kmem_cache_create("cifs_oplock_structs",
sizeof(struct oplock_q_entry), 0,
SLAB_HWCACHE_ALIGN, NULL);
if (cifs_oplock_cachep == NULL) {
mempool_destroy(cifs_mid_poolp);
kmem_cache_destroy(cifs_mid_cachep);
return -ENOMEM;
}
return 0;
}
static void
cifs_destroy_mids(void)
{
mempool_destroy(cifs_mid_poolp);
kmem_cache_destroy(cifs_mid_cachep);
kmem_cache_destroy(cifs_oplock_cachep);
}
static int cifs_oplock_thread(void *dummyarg)
{
struct oplock_q_entry *oplock_item;
struct cifsTconInfo *pTcon;
struct inode *inode;
__u16 netfid;
[CIFS] Fix potential data corruption when writing out cached dirty pages Fix RedHat bug 329431 The idea here is separate "conscious" from "unconscious" flushes. Conscious flushes are those due to a fsync() or close(). Unconscious ones are flushes that occur as a side effect of some other operation or due to memory pressure. Currently, when an error occurs during an unconscious flush (ENOSPC or EIO), we toss out the page and don't preserve that error to report to the user when a conscious flush occurs. If after the unconscious flush, there are no more dirty pages for the inode, the conscious flush will simply return success even though there were previous errors when writing out pages. This can lead to data corruption. The easiest way to reproduce this is to mount up a CIFS share that's very close to being full or where the user is very close to quota. mv a file to the share that's slightly larger than the quota allows. The writes will all succeed (since they go to pagecache). The mv will do a setattr to set the new file's attributes. This calls filemap_write_and_wait, which will return an error since all of the pages can't be written out. Then later, when the flush and release ops occur, there are no more dirty pages in pagecache for the file and those operations return 0. mv then assumes that the file was written out correctly and deletes the original. CIFS already has a write_behind_rc variable where it stores the results from earlier flushes, but that value is only reported in cifs_close. Since the VFS ignores the return value from the release operation, this isn't helpful. We should be reporting this error during the flush operation. This patch does the following: 1) changes cifs_fsync to use filemap_write_and_wait and cifs_flush and also sync to check its return code. If it returns successful, they then check the value of write_behind_rc to see if an earlier flush had reported any errors. If so, they return that error and clear write_behind_rc. 2) sets write_behind_rc in a few other places where pages are written out as a side effect of other operations and the code waits on them. 3) changes cifs_setattr to only call filemap_write_and_wait for ATTR_SIZE changes. 4) makes cifs_writepages accurately distinguish between EIO and ENOSPC errors when writing out pages. Some simple testing indicates that the patch works as expected and that it fixes the reproduceable known problem. Acked-by: Dave Kleikamp <shaggy@austin.rr.com> Signed-off-by: Jeff Layton <jlayton@redhat.com> Signed-off-by: Steve French <sfrench@us.ibm.com>
2007-11-21 06:19:03 +07:00
int rc, waitrc = 0;
set_freezable();
do {
if (try_to_freeze())
continue;
spin_lock(&GlobalMid_Lock);
if (list_empty(&GlobalOplock_Q)) {
spin_unlock(&GlobalMid_Lock);
set_current_state(TASK_INTERRUPTIBLE);
schedule_timeout(39*HZ);
} else {
oplock_item = list_entry(GlobalOplock_Q.next,
struct oplock_q_entry, qhead);
if (oplock_item) {
cFYI(1, ("found oplock item to write out"));
pTcon = oplock_item->tcon;
inode = oplock_item->pinode;
netfid = oplock_item->netfid;
spin_unlock(&GlobalMid_Lock);
DeleteOplockQEntry(oplock_item);
/* can not grab inode sem here since it would
deadlock when oplock received on delete
since vfs_unlink holds the i_mutex across
the call */
/* mutex_lock(&inode->i_mutex);*/
if (S_ISREG(inode->i_mode)) {
rc =
filemap_fdatawrite(inode->i_mapping);
if (CIFS_I(inode)->clientCanCacheRead
== 0) {
[CIFS] Fix potential data corruption when writing out cached dirty pages Fix RedHat bug 329431 The idea here is separate "conscious" from "unconscious" flushes. Conscious flushes are those due to a fsync() or close(). Unconscious ones are flushes that occur as a side effect of some other operation or due to memory pressure. Currently, when an error occurs during an unconscious flush (ENOSPC or EIO), we toss out the page and don't preserve that error to report to the user when a conscious flush occurs. If after the unconscious flush, there are no more dirty pages for the inode, the conscious flush will simply return success even though there were previous errors when writing out pages. This can lead to data corruption. The easiest way to reproduce this is to mount up a CIFS share that's very close to being full or where the user is very close to quota. mv a file to the share that's slightly larger than the quota allows. The writes will all succeed (since they go to pagecache). The mv will do a setattr to set the new file's attributes. This calls filemap_write_and_wait, which will return an error since all of the pages can't be written out. Then later, when the flush and release ops occur, there are no more dirty pages in pagecache for the file and those operations return 0. mv then assumes that the file was written out correctly and deletes the original. CIFS already has a write_behind_rc variable where it stores the results from earlier flushes, but that value is only reported in cifs_close. Since the VFS ignores the return value from the release operation, this isn't helpful. We should be reporting this error during the flush operation. This patch does the following: 1) changes cifs_fsync to use filemap_write_and_wait and cifs_flush and also sync to check its return code. If it returns successful, they then check the value of write_behind_rc to see if an earlier flush had reported any errors. If so, they return that error and clear write_behind_rc. 2) sets write_behind_rc in a few other places where pages are written out as a side effect of other operations and the code waits on them. 3) changes cifs_setattr to only call filemap_write_and_wait for ATTR_SIZE changes. 4) makes cifs_writepages accurately distinguish between EIO and ENOSPC errors when writing out pages. Some simple testing indicates that the patch works as expected and that it fixes the reproduceable known problem. Acked-by: Dave Kleikamp <shaggy@austin.rr.com> Signed-off-by: Jeff Layton <jlayton@redhat.com> Signed-off-by: Steve French <sfrench@us.ibm.com>
2007-11-21 06:19:03 +07:00
waitrc = filemap_fdatawait(inode->i_mapping);
invalidate_remote_inode(inode);
}
[CIFS] Fix potential data corruption when writing out cached dirty pages Fix RedHat bug 329431 The idea here is separate "conscious" from "unconscious" flushes. Conscious flushes are those due to a fsync() or close(). Unconscious ones are flushes that occur as a side effect of some other operation or due to memory pressure. Currently, when an error occurs during an unconscious flush (ENOSPC or EIO), we toss out the page and don't preserve that error to report to the user when a conscious flush occurs. If after the unconscious flush, there are no more dirty pages for the inode, the conscious flush will simply return success even though there were previous errors when writing out pages. This can lead to data corruption. The easiest way to reproduce this is to mount up a CIFS share that's very close to being full or where the user is very close to quota. mv a file to the share that's slightly larger than the quota allows. The writes will all succeed (since they go to pagecache). The mv will do a setattr to set the new file's attributes. This calls filemap_write_and_wait, which will return an error since all of the pages can't be written out. Then later, when the flush and release ops occur, there are no more dirty pages in pagecache for the file and those operations return 0. mv then assumes that the file was written out correctly and deletes the original. CIFS already has a write_behind_rc variable where it stores the results from earlier flushes, but that value is only reported in cifs_close. Since the VFS ignores the return value from the release operation, this isn't helpful. We should be reporting this error during the flush operation. This patch does the following: 1) changes cifs_fsync to use filemap_write_and_wait and cifs_flush and also sync to check its return code. If it returns successful, they then check the value of write_behind_rc to see if an earlier flush had reported any errors. If so, they return that error and clear write_behind_rc. 2) sets write_behind_rc in a few other places where pages are written out as a side effect of other operations and the code waits on them. 3) changes cifs_setattr to only call filemap_write_and_wait for ATTR_SIZE changes. 4) makes cifs_writepages accurately distinguish between EIO and ENOSPC errors when writing out pages. Some simple testing indicates that the patch works as expected and that it fixes the reproduceable known problem. Acked-by: Dave Kleikamp <shaggy@austin.rr.com> Signed-off-by: Jeff Layton <jlayton@redhat.com> Signed-off-by: Steve French <sfrench@us.ibm.com>
2007-11-21 06:19:03 +07:00
if (rc == 0)
rc = waitrc;
} else
rc = 0;
/* mutex_unlock(&inode->i_mutex);*/
if (rc)
CIFS_I(inode)->write_behind_rc = rc;
cFYI(1, ("Oplock flush inode %p rc %d",
inode, rc));
/* releasing stale oplock after recent reconnect
of smb session using a now incorrect file
handle is not a data integrity issue but do
not bother sending an oplock release if session
to server still is disconnected since oplock
already released by the server in that case */
if (pTcon->tidStatus != CifsNeedReconnect) {
rc = CIFSSMBLock(0, pTcon, netfid,
0 /* len */ , 0 /* offset */, 0,
0, LOCKING_ANDX_OPLOCK_RELEASE,
false /* wait flag */);
cFYI(1, ("Oplock release rc = %d", rc));
}
} else
spin_unlock(&GlobalMid_Lock);
set_current_state(TASK_INTERRUPTIBLE);
schedule_timeout(1); /* yield in case q were corrupt */
}
} while (!kthread_should_stop());
return 0;
}
static int cifs_dnotify_thread(void *dummyarg)
{
struct list_head *tmp;
struct cifsSesInfo *ses;
do {
if (try_to_freeze())
continue;
set_current_state(TASK_INTERRUPTIBLE);
schedule_timeout(15*HZ);
read_lock(&GlobalSMBSeslock);
/* check if any stuck requests that need
to be woken up and wakeq so the
thread can wake up and error out */
list_for_each(tmp, &GlobalSMBSessionList) {
ses = list_entry(tmp, struct cifsSesInfo,
cifsSessionList);
if (ses && ses->server &&
atomic_read(&ses->server->inFlight))
wake_up_all(&ses->server->response_q);
}
read_unlock(&GlobalSMBSeslock);
} while (!kthread_should_stop());
return 0;
}
static int __init
init_cifs(void)
{
int rc = 0;
cifs_proc_init();
/* INIT_LIST_HEAD(&GlobalServerList);*/ /* BB not implemented yet */
INIT_LIST_HEAD(&GlobalSMBSessionList);
INIT_LIST_HEAD(&GlobalTreeConnectionList);
INIT_LIST_HEAD(&GlobalOplock_Q);
#ifdef CONFIG_CIFS_EXPERIMENTAL
INIT_LIST_HEAD(&GlobalDnotifyReqList);
INIT_LIST_HEAD(&GlobalDnotifyRsp_Q);
#endif
/*
* Initialize Global counters
*/
atomic_set(&sesInfoAllocCount, 0);
atomic_set(&tconInfoAllocCount, 0);
atomic_set(&tcpSesAllocCount, 0);
atomic_set(&tcpSesReconnectCount, 0);
atomic_set(&tconInfoReconnectCount, 0);
atomic_set(&bufAllocCount, 0);
atomic_set(&smBufAllocCount, 0);
#ifdef CONFIG_CIFS_STATS2
atomic_set(&totBufAllocCount, 0);
atomic_set(&totSmBufAllocCount, 0);
#endif /* CONFIG_CIFS_STATS2 */
atomic_set(&midCount, 0);
GlobalCurrentXid = 0;
GlobalTotalActiveXid = 0;
GlobalMaxActiveXid = 0;
memset(Local_System_Name, 0, 15);
rwlock_init(&GlobalSMBSeslock);
spin_lock_init(&GlobalMid_Lock);
if (cifs_max_pending < 2) {
cifs_max_pending = 2;
cFYI(1, ("cifs_max_pending set to min of 2"));
} else if (cifs_max_pending > 256) {
cifs_max_pending = 256;
cFYI(1, ("cifs_max_pending set to max of 256"));
}
rc = cifs_init_inodecache();
if (rc)
goto out_clean_proc;
rc = cifs_init_mids();
if (rc)
goto out_destroy_inodecache;
rc = cifs_init_request_bufs();
if (rc)
goto out_destroy_mids;
rc = register_filesystem(&cifs_fs_type);
if (rc)
goto out_destroy_request_bufs;
#ifdef CONFIG_CIFS_UPCALL
rc = register_key_type(&cifs_spnego_key_type);
if (rc)
goto out_unregister_filesystem;
#endif
#ifdef CONFIG_CIFS_DFS_UPCALL
rc = register_key_type(&key_type_dns_resolver);
if (rc)
goto out_unregister_key_type;
#endif
oplockThread = kthread_run(cifs_oplock_thread, NULL, "cifsoplockd");
if (IS_ERR(oplockThread)) {
rc = PTR_ERR(oplockThread);
cERROR(1, ("error %d create oplock thread", rc));
goto out_unregister_dfs_key_type;
}
dnotifyThread = kthread_run(cifs_dnotify_thread, NULL, "cifsdnotifyd");
if (IS_ERR(dnotifyThread)) {
rc = PTR_ERR(dnotifyThread);
cERROR(1, ("error %d create dnotify thread", rc));
goto out_stop_oplock_thread;
}
return 0;
out_stop_oplock_thread:
kthread_stop(oplockThread);
out_unregister_dfs_key_type:
#ifdef CONFIG_CIFS_DFS_UPCALL
unregister_key_type(&key_type_dns_resolver);
out_unregister_key_type:
#endif
#ifdef CONFIG_CIFS_UPCALL
unregister_key_type(&cifs_spnego_key_type);
out_unregister_filesystem:
#endif
unregister_filesystem(&cifs_fs_type);
out_destroy_request_bufs:
cifs_destroy_request_bufs();
out_destroy_mids:
cifs_destroy_mids();
out_destroy_inodecache:
cifs_destroy_inodecache();
out_clean_proc:
cifs_proc_clean();
return rc;
}
static void __exit
exit_cifs(void)
{
cFYI(DBG2, ("exit_cifs"));
cifs_proc_clean();
#ifdef CONFIG_CIFS_DFS_UPCALL
cifs_dfs_release_automount_timer();
unregister_key_type(&key_type_dns_resolver);
#endif
#ifdef CONFIG_CIFS_UPCALL
unregister_key_type(&cifs_spnego_key_type);
#endif
unregister_filesystem(&cifs_fs_type);
cifs_destroy_inodecache();
cifs_destroy_mids();
cifs_destroy_request_bufs();
kthread_stop(oplockThread);
kthread_stop(dnotifyThread);
}
MODULE_AUTHOR("Steve French <sfrench@us.ibm.com>");
MODULE_LICENSE("GPL"); /* combination of LGPL + GPL source behaves as GPL */
MODULE_DESCRIPTION
("VFS to access servers complying with the SNIA CIFS Specification "
"e.g. Samba and Windows");
MODULE_VERSION(CIFS_VERSION);
module_init(init_cifs)
module_exit(exit_cifs)