linux-stable/fs/vboxsf/dir.c

486 lines
12 KiB
C
Raw Normal View History

// SPDX-License-Identifier: MIT
/*
* VirtualBox Guest Shared Folders support: Directory inode and file operations
*
* Copyright (C) 2006-2018 Oracle Corporation
*/
#include <linux/namei.h>
#include <linux/vbox_utils.h>
#include "vfsmod.h"
static int vboxsf_dir_open(struct inode *inode, struct file *file)
{
struct vboxsf_sbi *sbi = VBOXSF_SBI(inode->i_sb);
struct shfl_createparms params = {};
struct vboxsf_dir_info *sf_d;
int err;
sf_d = vboxsf_dir_info_alloc();
if (!sf_d)
return -ENOMEM;
params.handle = SHFL_HANDLE_NIL;
params.create_flags = SHFL_CF_DIRECTORY | SHFL_CF_ACT_OPEN_IF_EXISTS |
SHFL_CF_ACT_FAIL_IF_NEW | SHFL_CF_ACCESS_READ;
err = vboxsf_create_at_dentry(file_dentry(file), &params);
if (err)
goto err_free_dir_info;
if (params.result != SHFL_FILE_EXISTS) {
err = -ENOENT;
goto err_close;
}
err = vboxsf_dir_read_all(sbi, sf_d, params.handle);
if (err)
goto err_close;
vboxsf_close(sbi->root, params.handle);
file->private_data = sf_d;
return 0;
err_close:
vboxsf_close(sbi->root, params.handle);
err_free_dir_info:
vboxsf_dir_info_free(sf_d);
return err;
}
static int vboxsf_dir_release(struct inode *inode, struct file *file)
{
if (file->private_data)
vboxsf_dir_info_free(file->private_data);
return 0;
}
static unsigned int vboxsf_get_d_type(u32 mode)
{
unsigned int d_type;
switch (mode & SHFL_TYPE_MASK) {
case SHFL_TYPE_FIFO:
d_type = DT_FIFO;
break;
case SHFL_TYPE_DEV_CHAR:
d_type = DT_CHR;
break;
case SHFL_TYPE_DIRECTORY:
d_type = DT_DIR;
break;
case SHFL_TYPE_DEV_BLOCK:
d_type = DT_BLK;
break;
case SHFL_TYPE_FILE:
d_type = DT_REG;
break;
case SHFL_TYPE_SYMLINK:
d_type = DT_LNK;
break;
case SHFL_TYPE_SOCKET:
d_type = DT_SOCK;
break;
case SHFL_TYPE_WHITEOUT:
d_type = DT_WHT;
break;
default:
d_type = DT_UNKNOWN;
break;
}
return d_type;
}
static bool vboxsf_dir_emit(struct file *dir, struct dir_context *ctx)
{
struct vboxsf_sbi *sbi = VBOXSF_SBI(file_inode(dir)->i_sb);
struct vboxsf_dir_info *sf_d = dir->private_data;
struct shfl_dirinfo *info;
struct vboxsf_dir_buf *b;
unsigned int d_type;
loff_t i, cur = 0;
ino_t fake_ino;
void *end;
int err;
list_for_each_entry(b, &sf_d->info_list, head) {
try_next_entry:
if (ctx->pos >= cur + b->entries) {
cur += b->entries;
continue;
}
/*
* Note the vboxsf_dir_info objects we are iterating over here
* are variable sized, so the info pointer may end up being
* unaligned. This is how we get the data from the host.
* Since vboxsf is only supported on x86 machines this is not
* a problem.
*/
for (i = 0, info = b->buf; i < ctx->pos - cur; i++) {
end = &info->name.string.utf8[info->name.size];
/* Only happens if the host gives us corrupt data */
if (WARN_ON(end > (b->buf + b->used)))
return false;
info = end;
}
end = &info->name.string.utf8[info->name.size];
if (WARN_ON(end > (b->buf + b->used)))
return false;
/* Info now points to the right entry, emit it. */
d_type = vboxsf_get_d_type(info->info.attr.mode);
/*
* On 32-bit systems pos is 64-bit signed, while ino is 32-bit
* unsigned so fake_ino may overflow, check for this.
*/
if ((ino_t)(ctx->pos + 1) != (u64)(ctx->pos + 1)) {
vbg_err("vboxsf: fake ino overflow, truncating dir\n");
return false;
}
fake_ino = ctx->pos + 1;
if (sbi->nls) {
char d_name[NAME_MAX];
err = vboxsf_nlscpy(sbi, d_name, NAME_MAX,
info->name.string.utf8,
info->name.length);
if (err) {
/* skip erroneous entry and proceed */
ctx->pos += 1;
goto try_next_entry;
}
return dir_emit(ctx, d_name, strlen(d_name),
fake_ino, d_type);
}
return dir_emit(ctx, info->name.string.utf8, info->name.length,
fake_ino, d_type);
}
return false;
}
static int vboxsf_dir_iterate(struct file *dir, struct dir_context *ctx)
{
bool emitted;
do {
emitted = vboxsf_dir_emit(dir, ctx);
if (emitted)
ctx->pos += 1;
} while (emitted);
return 0;
}
const struct file_operations vboxsf_dir_fops = {
.open = vboxsf_dir_open,
.iterate = vboxsf_dir_iterate,
.release = vboxsf_dir_release,
.read = generic_read_dir,
.llseek = generic_file_llseek,
};
/*
* This is called during name resolution/lookup to check if the @dentry in
* the cache is still valid. the job is handled by vboxsf_inode_revalidate.
*/
static int vboxsf_dentry_revalidate(struct dentry *dentry, unsigned int flags)
{
if (flags & LOOKUP_RCU)
return -ECHILD;
if (d_really_is_positive(dentry))
return vboxsf_inode_revalidate(dentry) == 0;
else
return vboxsf_stat_dentry(dentry, NULL) == -ENOENT;
}
const struct dentry_operations vboxsf_dentry_ops = {
.d_revalidate = vboxsf_dentry_revalidate
};
/* iops */
static struct dentry *vboxsf_dir_lookup(struct inode *parent,
struct dentry *dentry,
unsigned int flags)
{
struct vboxsf_sbi *sbi = VBOXSF_SBI(parent->i_sb);
struct shfl_fsobjinfo fsinfo;
struct inode *inode;
int err;
dentry->d_time = jiffies;
err = vboxsf_stat_dentry(dentry, &fsinfo);
if (err) {
inode = (err == -ENOENT) ? NULL : ERR_PTR(err);
} else {
inode = vboxsf_new_inode(parent->i_sb);
if (!IS_ERR(inode))
vboxsf_init_inode(sbi, inode, &fsinfo, false);
}
return d_splice_alias(inode, dentry);
}
static int vboxsf_dir_instantiate(struct inode *parent, struct dentry *dentry,
struct shfl_fsobjinfo *info)
{
struct vboxsf_sbi *sbi = VBOXSF_SBI(parent->i_sb);
struct vboxsf_inode *sf_i;
struct inode *inode;
inode = vboxsf_new_inode(parent->i_sb);
if (IS_ERR(inode))
return PTR_ERR(inode);
sf_i = VBOXSF_I(inode);
/* The host may have given us different attr then requested */
sf_i->force_restat = 1;
vboxsf_init_inode(sbi, inode, info, false);
d_instantiate(dentry, inode);
return 0;
}
static int vboxsf_dir_create(struct inode *parent, struct dentry *dentry,
umode_t mode, bool is_dir, bool excl, u64 *handle_ret)
{
struct vboxsf_inode *sf_parent_i = VBOXSF_I(parent);
struct vboxsf_sbi *sbi = VBOXSF_SBI(parent->i_sb);
struct shfl_createparms params = {};
int err;
params.handle = SHFL_HANDLE_NIL;
params.create_flags = SHFL_CF_ACT_CREATE_IF_NEW | SHFL_CF_ACCESS_READWRITE;
if (is_dir)
params.create_flags |= SHFL_CF_DIRECTORY;
if (excl)
params.create_flags |= SHFL_CF_ACT_FAIL_IF_EXISTS;
params.info.attr.mode = (mode & 0777) |
(is_dir ? SHFL_TYPE_DIRECTORY : SHFL_TYPE_FILE);
params.info.attr.additional = SHFLFSOBJATTRADD_NOTHING;
err = vboxsf_create_at_dentry(dentry, &params);
if (err)
return err;
if (params.result != SHFL_FILE_CREATED)
return -EPERM;
err = vboxsf_dir_instantiate(parent, dentry, &params.info);
if (err)
goto out;
/* parent directory access/change time changed */
sf_parent_i->force_restat = 1;
out:
if (err == 0 && handle_ret)
*handle_ret = params.handle;
else
vboxsf_close(sbi->root, params.handle);
return err;
}
static int vboxsf_dir_mkfile(struct user_namespace *mnt_userns,
struct inode *parent, struct dentry *dentry,
umode_t mode, bool excl)
{
return vboxsf_dir_create(parent, dentry, mode, false, excl, NULL);
}
static int vboxsf_dir_mkdir(struct user_namespace *mnt_userns,
struct inode *parent, struct dentry *dentry,
umode_t mode)
{
return vboxsf_dir_create(parent, dentry, mode, true, true, NULL);
}
vboxsf: Add support for the atomic_open directory-inode op Opening a new file is done in 2 steps on regular filesystems: 1. Call the create inode-op on the parent-dir to create an inode to hold the meta-data related to the file. 2. Call the open file-op to get a handle for the file. vboxsf however does not really use disk-backed inodes because it is based on passing through file-related system-calls through to the hypervisor. So both steps translate to an open(2) call being passed through to the hypervisor. With the handle returned by the first call immediately being closed again. Making 2 open calls for a single open(..., O_CREATE, ...) calls has 2 problems: a) It is not really efficient. b) It actually breaks some apps. An example of b) is doing a git clone inside a vboxsf mount. When git clone tries to create a tempfile to store the pak files which is downloading the following happens: 1. vboxsf_dir_mkfile() gets called with a mode of 0444 and succeeds. 2. vboxsf_file_open() gets called with file->f_flags containing O_RDWR. When the host is a Linux machine this fails because doing a open(..., O_RDWR) on a file which exists and has mode 0444 results in an -EPERM error. Other network-filesystems and fuse avoid the problem of needing to pass 2 open() calls to the other side by using the atomic_open directory-inode op. This commit fixes git clone not working inside a vboxsf mount, by adding support for the atomic_open directory-inode op. As an added bonus this should also make opening new files faster. The atomic_open implementation is modelled after the atomic_open implementations from the 9p and fuse code. Fixes: 0fd169576648 ("fs: Add VirtualBox guest shared folder (vboxsf) support") Reported-by: Ludovic Pouzenc <bugreports@pouzenc.fr> Signed-off-by: Hans de Goede <hdegoede@redhat.com>
2021-01-21 11:54:18 +00:00
static int vboxsf_dir_atomic_open(struct inode *parent, struct dentry *dentry,
struct file *file, unsigned int flags, umode_t mode)
{
struct vboxsf_sbi *sbi = VBOXSF_SBI(parent->i_sb);
struct vboxsf_handle *sf_handle;
struct dentry *res = NULL;
u64 handle;
int err;
if (d_in_lookup(dentry)) {
res = vboxsf_dir_lookup(parent, dentry, 0);
if (IS_ERR(res))
return PTR_ERR(res);
if (res)
dentry = res;
}
/* Only creates */
if (!(flags & O_CREAT) || d_really_is_positive(dentry))
return finish_no_open(file, res);
err = vboxsf_dir_create(parent, dentry, mode, false, flags & O_EXCL, &handle);
if (err)
goto out;
sf_handle = vboxsf_create_sf_handle(d_inode(dentry), handle, SHFL_CF_ACCESS_READWRITE);
if (IS_ERR(sf_handle)) {
vboxsf_close(sbi->root, handle);
err = PTR_ERR(sf_handle);
goto out;
}
err = finish_open(file, dentry, generic_file_open);
if (err) {
/* This also closes the handle passed to vboxsf_create_sf_handle() */
vboxsf_release_sf_handle(d_inode(dentry), sf_handle);
goto out;
}
file->private_data = sf_handle;
file->f_mode |= FMODE_CREATED;
out:
dput(res);
return err;
}
static int vboxsf_dir_unlink(struct inode *parent, struct dentry *dentry)
{
struct vboxsf_sbi *sbi = VBOXSF_SBI(parent->i_sb);
struct vboxsf_inode *sf_parent_i = VBOXSF_I(parent);
struct inode *inode = d_inode(dentry);
struct shfl_string *path;
u32 flags;
int err;
if (S_ISDIR(inode->i_mode))
flags = SHFL_REMOVE_DIR;
else
flags = SHFL_REMOVE_FILE;
if (S_ISLNK(inode->i_mode))
flags |= SHFL_REMOVE_SYMLINK;
path = vboxsf_path_from_dentry(sbi, dentry);
if (IS_ERR(path))
return PTR_ERR(path);
err = vboxsf_remove(sbi->root, path, flags);
__putname(path);
if (err)
return err;
/* parent directory access/change time changed */
sf_parent_i->force_restat = 1;
return 0;
}
static int vboxsf_dir_rename(struct user_namespace *mnt_userns,
struct inode *old_parent,
struct dentry *old_dentry,
struct inode *new_parent,
struct dentry *new_dentry,
unsigned int flags)
{
struct vboxsf_sbi *sbi = VBOXSF_SBI(old_parent->i_sb);
struct vboxsf_inode *sf_old_parent_i = VBOXSF_I(old_parent);
struct vboxsf_inode *sf_new_parent_i = VBOXSF_I(new_parent);
u32 shfl_flags = SHFL_RENAME_FILE | SHFL_RENAME_REPLACE_IF_EXISTS;
struct shfl_string *old_path, *new_path;
int err;
if (flags)
return -EINVAL;
old_path = vboxsf_path_from_dentry(sbi, old_dentry);
if (IS_ERR(old_path))
return PTR_ERR(old_path);
new_path = vboxsf_path_from_dentry(sbi, new_dentry);
if (IS_ERR(new_path)) {
err = PTR_ERR(new_path);
goto err_put_old_path;
}
if (d_inode(old_dentry)->i_mode & S_IFDIR)
shfl_flags = 0;
err = vboxsf_rename(sbi->root, old_path, new_path, shfl_flags);
if (err == 0) {
/* parent directories access/change time changed */
sf_new_parent_i->force_restat = 1;
sf_old_parent_i->force_restat = 1;
}
__putname(new_path);
err_put_old_path:
__putname(old_path);
return err;
}
static int vboxsf_dir_symlink(struct user_namespace *mnt_userns,
struct inode *parent, struct dentry *dentry,
const char *symname)
{
struct vboxsf_inode *sf_parent_i = VBOXSF_I(parent);
struct vboxsf_sbi *sbi = VBOXSF_SBI(parent->i_sb);
int symname_size = strlen(symname) + 1;
struct shfl_string *path, *ssymname;
struct shfl_fsobjinfo info;
int err;
path = vboxsf_path_from_dentry(sbi, dentry);
if (IS_ERR(path))
return PTR_ERR(path);
ssymname = kmalloc(SHFLSTRING_HEADER_SIZE + symname_size, GFP_KERNEL);
if (!ssymname) {
__putname(path);
return -ENOMEM;
}
ssymname->length = symname_size - 1;
ssymname->size = symname_size;
memcpy(ssymname->string.utf8, symname, symname_size);
err = vboxsf_symlink(sbi->root, path, ssymname, &info);
kfree(ssymname);
__putname(path);
if (err) {
/* -EROFS means symlinks are note support -> -EPERM */
return (err == -EROFS) ? -EPERM : err;
}
err = vboxsf_dir_instantiate(parent, dentry, &info);
if (err)
return err;
/* parent directory access/change time changed */
sf_parent_i->force_restat = 1;
return 0;
}
const struct inode_operations vboxsf_dir_iops = {
.lookup = vboxsf_dir_lookup,
.create = vboxsf_dir_mkfile,
.mkdir = vboxsf_dir_mkdir,
vboxsf: Add support for the atomic_open directory-inode op Opening a new file is done in 2 steps on regular filesystems: 1. Call the create inode-op on the parent-dir to create an inode to hold the meta-data related to the file. 2. Call the open file-op to get a handle for the file. vboxsf however does not really use disk-backed inodes because it is based on passing through file-related system-calls through to the hypervisor. So both steps translate to an open(2) call being passed through to the hypervisor. With the handle returned by the first call immediately being closed again. Making 2 open calls for a single open(..., O_CREATE, ...) calls has 2 problems: a) It is not really efficient. b) It actually breaks some apps. An example of b) is doing a git clone inside a vboxsf mount. When git clone tries to create a tempfile to store the pak files which is downloading the following happens: 1. vboxsf_dir_mkfile() gets called with a mode of 0444 and succeeds. 2. vboxsf_file_open() gets called with file->f_flags containing O_RDWR. When the host is a Linux machine this fails because doing a open(..., O_RDWR) on a file which exists and has mode 0444 results in an -EPERM error. Other network-filesystems and fuse avoid the problem of needing to pass 2 open() calls to the other side by using the atomic_open directory-inode op. This commit fixes git clone not working inside a vboxsf mount, by adding support for the atomic_open directory-inode op. As an added bonus this should also make opening new files faster. The atomic_open implementation is modelled after the atomic_open implementations from the 9p and fuse code. Fixes: 0fd169576648 ("fs: Add VirtualBox guest shared folder (vboxsf) support") Reported-by: Ludovic Pouzenc <bugreports@pouzenc.fr> Signed-off-by: Hans de Goede <hdegoede@redhat.com>
2021-01-21 11:54:18 +00:00
.atomic_open = vboxsf_dir_atomic_open,
.rmdir = vboxsf_dir_unlink,
.unlink = vboxsf_dir_unlink,
.rename = vboxsf_dir_rename,
.symlink = vboxsf_dir_symlink,
.getattr = vboxsf_getattr,
.setattr = vboxsf_setattr,
};