/*
* Remount as initiated by pvfs2-client-core on restart. This is used to
* repopulate mount information left from previous pvfs2-client-core.
*
* the idea here is that given a valid superblock, we're
* re-initializing the user space client with the initial mount
* information specified when the super block was first initialized.
* this is very different than the first initialization/creation of a
* superblock. we use the special service_priority_operation to make
* sure that the mount gets ahead of any other pending operation that
* is waiting for servicing. this means that the pvfs2-client won't
* fail to start several times for all other pending operations before
* the client regains all of the mount information from us.
* NOTE: this function assumes that the request_mutex is already acquired!
*/
int orangefs_remount(struct orangefs_sb_info_s *orangefs_sb)
{
struct orangefs_kernel_op_s *new_op;
int ret = -EINVAL;
gossip_debug(GOSSIP_SUPER_DEBUG, "orangefs_remount: called\n");
new_op = op_alloc(ORANGEFS_VFS_OP_FS_MOUNT);
if (!new_op)
return -ENOMEM;
strncpy(new_op->upcall.req.fs_mount.orangefs_config_server,
orangefs_sb->devname,
ORANGEFS_MAX_SERVER_ADDR_LEN);
gossip_debug(GOSSIP_SUPER_DEBUG,
"Attempting ORANGEFS Remount via host %s\n",
new_op->upcall.req.fs_mount.orangefs_config_server);
/*
* we assume that the calling function has already acquired the
* request_mutex to prevent other operations from bypassing
* this one
*/
ret = service_operation(new_op, "orangefs_remount",
ORANGEFS_OP_PRIORITY | ORANGEFS_OP_NO_MUTEX);
gossip_debug(GOSSIP_SUPER_DEBUG,
"orangefs_remount: mount got return value of %d\n",
ret);
if (ret == 0) {
/*
* store the id assigned to this sb -- it's just a
* short-lived mapping that the system interface uses
* to map this superblock to a particular mount entry
*/
orangefs_sb->id = new_op->downcall.resp.fs_mount.id;
orangefs_sb->mount_pending = 0;
}
op_release(new_op);
if (orangefs_userspace_version >= 20906) {
new_op = op_alloc(ORANGEFS_VFS_OP_FEATURES);
if (!new_op)
return -ENOMEM;
new_op->upcall.req.features.features = 0;
ret = service_operation(new_op, "orangefs_features",
ORANGEFS_OP_PRIORITY | ORANGEFS_OP_NO_MUTEX);
if (!ret)
orangefs_features =
new_op->downcall.resp.features.features;
else
orangefs_features = 0;
op_release(new_op);
} else {
orangefs_features = 0;
}
return ret;
}
static int flush_racache(struct inode *inode)
{
struct orangefs_inode_s *orangefs_inode = ORANGEFS_I(inode);
struct orangefs_kernel_op_s *new_op;
int ret;
gossip_debug(GOSSIP_UTILS_DEBUG,
"%s: %pU: Handle is %pU | fs_id %d\n", __func__,
get_khandle_from_ino(inode), &orangefs_inode->refn.khandle,
orangefs_inode->refn.fs_id);
new_op = op_alloc(ORANGEFS_VFS_OP_RA_FLUSH);
if (!new_op)
return -ENOMEM;
new_op->upcall.req.ra_cache_flush.refn = orangefs_inode->refn;
ret = service_operation(new_op, "orangefs_flush_racache",
get_interruptible_flag(inode));
gossip_debug(GOSSIP_UTILS_DEBUG, "%s: got return value of %d\n",
__func__, ret);
op_release(new_op);
return ret;
}
/*
* Push all data for a specific file onto permanent storage.
*/
static int orangefs_fsync(struct file *file,
loff_t start,
loff_t end,
int datasync)
{
int ret = -EINVAL;
struct orangefs_inode_s *orangefs_inode =
ORANGEFS_I(file->f_path.dentry->d_inode);
struct orangefs_kernel_op_s *new_op = NULL;
/* required call */
filemap_write_and_wait_range(file->f_mapping, start, end);
new_op = op_alloc(ORANGEFS_VFS_OP_FSYNC);
if (!new_op)
return -ENOMEM;
new_op->upcall.req.fsync.refn = orangefs_inode->refn;
ret = service_operation(new_op,
"orangefs_fsync",
get_interruptible_flag(file->f_path.dentry->d_inode));
gossip_debug(GOSSIP_FILE_DEBUG,
"orangefs_fsync got return value of %d\n",
ret);
op_release(new_op);
orangefs_flush_inode(file->f_path.dentry->d_inode);
return ret;
}
/*
* Push all data for a specific file onto permanent storage.
*/
static int orangefs_fsync(struct file *file,
loff_t start,
loff_t end,
int datasync)
{
int ret;
struct orangefs_inode_s *orangefs_inode =
ORANGEFS_I(file_inode(file));
struct orangefs_kernel_op_s *new_op = NULL;
new_op = op_alloc(ORANGEFS_VFS_OP_FSYNC);
if (!new_op)
return -ENOMEM;
new_op->upcall.req.fsync.refn = orangefs_inode->refn;
ret = service_operation(new_op,
"orangefs_fsync",
get_interruptible_flag(file_inode(file)));
gossip_debug(GOSSIP_FILE_DEBUG,
"orangefs_fsync got return value of %d\n",
ret);
op_release(new_op);
orangefs_flush_inode(file_inode(file));
return ret;
}
static int orangefs_setattr_size(struct inode *inode, struct iattr *iattr)
{
struct orangefs_inode_s *orangefs_inode = ORANGEFS_I(inode);
struct orangefs_kernel_op_s *new_op;
loff_t orig_size;
int ret = -EINVAL;
gossip_debug(GOSSIP_INODE_DEBUG,
"%s: %pU: Handle is %pU | fs_id %d | size is %llu\n",
__func__,
get_khandle_from_ino(inode),
&orangefs_inode->refn.khandle,
orangefs_inode->refn.fs_id,
iattr->ia_size);
/* Ensure that we have a up to date size, so we know if it changed. */
ret = orangefs_inode_getattr(inode, 0, 1);
if (ret == -ESTALE)
ret = -EIO;
if (ret) {
gossip_err("%s: orangefs_inode_getattr failed, ret:%d:.\n",
__func__, ret);
return ret;
}
orig_size = i_size_read(inode);
truncate_setsize(inode, iattr->ia_size);
new_op = op_alloc(ORANGEFS_VFS_OP_TRUNCATE);
if (!new_op)
return -ENOMEM;
new_op->upcall.req.truncate.refn = orangefs_inode->refn;
new_op->upcall.req.truncate.size = (__s64) iattr->ia_size;
ret = service_operation(new_op, __func__,
get_interruptible_flag(inode));
/*
* the truncate has no downcall members to retrieve, but
* the status value tells us if it went through ok or not
*/
gossip_debug(GOSSIP_INODE_DEBUG,
"orangefs: orangefs_truncate got return value of %d\n",
ret);
op_release(new_op);
if (ret != 0)
return ret;
if (orig_size != i_size_read(inode))
iattr->ia_valid |= ATTR_CTIME | ATTR_MTIME;
return ret;
}
/*
* NOTE: information filled in here is typically reflected in the
* output of the system command 'df'
*/
static int orangefs_statfs(struct dentry *dentry, struct kstatfs *buf)
{
int ret = -ENOMEM;
struct orangefs_kernel_op_s *new_op = NULL;
int flags = 0;
struct super_block *sb = NULL;
sb = dentry->d_sb;
gossip_debug(GOSSIP_SUPER_DEBUG,
"orangefs_statfs: called on sb %p (fs_id is %d)\n",
sb,
(int)(ORANGEFS_SB(sb)->fs_id));
new_op = op_alloc(ORANGEFS_VFS_OP_STATFS);
if (!new_op)
return ret;
new_op->upcall.req.statfs.fs_id = ORANGEFS_SB(sb)->fs_id;
if (ORANGEFS_SB(sb)->flags & ORANGEFS_OPT_INTR)
flags = ORANGEFS_OP_INTERRUPTIBLE;
ret = service_operation(new_op, "orangefs_statfs", flags);
if (new_op->downcall.status < 0)
goto out_op_release;
gossip_debug(GOSSIP_SUPER_DEBUG,
"%s: got %ld blocks available | "
"%ld blocks total | %ld block size | "
"%ld files total | %ld files avail\n",
__func__,
(long)new_op->downcall.resp.statfs.blocks_avail,
(long)new_op->downcall.resp.statfs.blocks_total,
(long)new_op->downcall.resp.statfs.block_size,
(long)new_op->downcall.resp.statfs.files_total,
(long)new_op->downcall.resp.statfs.files_avail);
buf->f_type = sb->s_magic;
memcpy(&buf->f_fsid, &ORANGEFS_SB(sb)->fs_id, sizeof(buf->f_fsid));
buf->f_bsize = new_op->downcall.resp.statfs.block_size;
buf->f_namelen = ORANGEFS_NAME_MAX;
buf->f_blocks = (sector_t) new_op->downcall.resp.statfs.blocks_total;
buf->f_bfree = (sector_t) new_op->downcall.resp.statfs.blocks_avail;
buf->f_bavail = (sector_t) new_op->downcall.resp.statfs.blocks_avail;
buf->f_files = (sector_t) new_op->downcall.resp.statfs.files_total;
buf->f_ffree = (sector_t) new_op->downcall.resp.statfs.files_avail;
buf->f_frsize = sb->s_blocksize;
out_op_release:
op_release(new_op);
gossip_debug(GOSSIP_SUPER_DEBUG, "orangefs_statfs: returning %d\n", ret);
return ret;
}
static int orangefs_inode_removexattr(struct inode *inode,
const char *prefix,
const char *name,
int flags)
{
struct orangefs_inode_s *orangefs_inode = ORANGEFS_I(inode);
struct orangefs_kernel_op_s *new_op = NULL;
int ret = -ENOMEM;
down_write(&orangefs_inode->xattr_sem);
new_op = op_alloc(ORANGEFS_VFS_OP_REMOVEXATTR);
if (!new_op)
goto out_unlock;
new_op->upcall.req.removexattr.refn = orangefs_inode->refn;
/*
* NOTE: Although keys are meant to be NULL terminated
* textual strings, I am going to explicitly pass the
* length just in case we change this later on...
*/
ret = snprintf((char *)new_op->upcall.req.removexattr.key,
ORANGEFS_MAX_XATTR_NAMELEN,
"%s%s",
(prefix ? prefix : ""),
name);
new_op->upcall.req.removexattr.key_sz = ret + 1;
gossip_debug(GOSSIP_XATTR_DEBUG,
"orangefs_inode_removexattr: key %s, key_sz %d\n",
(char *)new_op->upcall.req.removexattr.key,
(int)new_op->upcall.req.removexattr.key_sz);
ret = service_operation(new_op,
"orangefs_inode_removexattr",
get_interruptible_flag(inode));
if (ret == -ENOENT) {
/*
* Request to replace a non-existent attribute is an error.
*/
if (flags & XATTR_REPLACE)
ret = -ENODATA;
else
ret = 0;
}
gossip_debug(GOSSIP_XATTR_DEBUG,
"orangefs_inode_removexattr: returning %d\n", ret);
op_release(new_op);
out_unlock:
up_write(&orangefs_inode->xattr_sem);
return ret;
}
static int orangefs_unmount(int id, __s32 fs_id, const char *devname)
{
struct orangefs_kernel_op_s *op;
int r;
op = op_alloc(ORANGEFS_VFS_OP_FS_UMOUNT);
if (!op)
return -ENOMEM;
op->upcall.req.fs_umount.id = id;
op->upcall.req.fs_umount.fs_id = fs_id;
strncpy(op->upcall.req.fs_umount.orangefs_config_server,
devname, ORANGEFS_MAX_SERVER_ADDR_LEN - 1);
r = service_operation(op, "orangefs_fs_umount", 0);
/* Not much to do about an error here. */
if (r)
gossip_err("orangefs_unmount: service_operation %d\n", r);
op_release(op);
return r;
}
void Interpreter::step()
{
if (instruction >= file->getCodeWordCount()) return;
if (system->getTick() < waitUntil)
return;
const uint16_t *code = file->getCode();
if (code[instruction] & (1 << 11))
{
// Short format
unsigned opcode = (code[instruction] & 0x0700) >> 8;
int argDiff = (int8_t) (code[instruction] & 0xFF);
uint16_t arguments[2];
switch(opcode)
{
case 0: // SHORT_OP_MOV
arguments[0] = argDiff + code[instruction + 1];
arguments[1] = code[instruction + 1];
instruction += 2;
op_mov(0, arguments);
break;
case 1: // SHORT_OP_ACQUIRE
arguments[0] = argDiff;
instruction += 1;
op_acquire(0, arguments);
break;
case 2: // SHORT_OP_RELEASE
arguments[0] = argDiff;
instruction += 1;
op_release(0, arguments);
break;
case 3: // SHORT_OP_SUBCALL
arguments[0] = argDiff + code[instruction + 1]; // Subroutine
arguments[1] = code[instruction + 1]; // Caller ID
instruction += 2; // Instruction has to be on the next before
// branching to subroutine so it can be used as return inst.
op_subcall(0, arguments);
break;
default:
throw std::runtime_error("Invalid short opcode.");
}
}
/* return 0 on success; non-zero otherwise */
static int orangefs_unlink(struct inode *dir, struct dentry *dentry)
{
struct inode *inode = dentry->d_inode;
struct orangefs_inode_s *parent = ORANGEFS_I(dir);
struct orangefs_kernel_op_s *new_op;
int ret;
gossip_debug(GOSSIP_NAME_DEBUG,
"%s: called on %s\n"
" (inode %pU): Parent is %pU | fs_id %d\n",
__func__,
dentry->d_name.name,
get_khandle_from_ino(inode),
&parent->refn.khandle,
parent->refn.fs_id);
new_op = op_alloc(ORANGEFS_VFS_OP_REMOVE);
if (!new_op)
return -ENOMEM;
new_op->upcall.req.remove.parent_refn = parent->refn;
strncpy(new_op->upcall.req.remove.d_name, dentry->d_name.name,
ORANGEFS_NAME_MAX);
ret = service_operation(new_op, "orangefs_unlink",
get_interruptible_flag(inode));
gossip_debug(GOSSIP_NAME_DEBUG,
"%s: service_operation returned:%d:\n",
__func__,
ret);
op_release(new_op);
if (!ret) {
drop_nlink(inode);
SetMtimeFlag(parent);
dir->i_mtime = dir->i_ctime = current_fs_time(dir->i_sb);
mark_inode_dirty_sync(dir);
}
return ret;
}
static int orangefs_rename(struct inode *old_dir,
struct dentry *old_dentry,
struct inode *new_dir,
struct dentry *new_dentry)
{
struct orangefs_kernel_op_s *new_op;
int ret;
gossip_debug(GOSSIP_NAME_DEBUG,
"orangefs_rename: called (%pd2 => %pd2) ct=%d\n",
old_dentry, new_dentry, d_count(new_dentry));
ORANGEFS_I(new_dentry->d_parent->d_inode)->getattr_time = jiffies - 1;
new_op = op_alloc(ORANGEFS_VFS_OP_RENAME);
if (!new_op)
return -EINVAL;
new_op->upcall.req.rename.old_parent_refn = ORANGEFS_I(old_dir)->refn;
new_op->upcall.req.rename.new_parent_refn = ORANGEFS_I(new_dir)->refn;
strncpy(new_op->upcall.req.rename.d_old_name,
old_dentry->d_name.name,
ORANGEFS_NAME_MAX);
strncpy(new_op->upcall.req.rename.d_new_name,
new_dentry->d_name.name,
ORANGEFS_NAME_MAX);
ret = service_operation(new_op,
"orangefs_rename",
get_interruptible_flag(old_dentry->d_inode));
gossip_debug(GOSSIP_NAME_DEBUG,
"orangefs_rename: got downcall status %d\n",
ret);
if (new_dentry->d_inode)
new_dentry->d_inode->i_ctime = CURRENT_TIME;
op_release(new_op);
return ret;
}
/*
* Attempt to resolve an object name (dentry->d_name), parent handle, and
* fsid into a handle for the object.
*/
static struct dentry *orangefs_lookup(struct inode *dir, struct dentry *dentry,
unsigned int flags)
{
struct orangefs_inode_s *parent = ORANGEFS_I(dir);
struct orangefs_kernel_op_s *new_op;
struct inode *inode;
struct dentry *res;
int ret = -EINVAL;
/*
* in theory we could skip a lookup here (if the intent is to
* create) in order to avoid a potentially failed lookup, but
* leaving it in can skip a valid lookup and try to create a file
* that already exists (e.g. the vfs already handles checking for
* -EEXIST on O_EXCL opens, which is broken if we skip this lookup
* in the create path)
*/
gossip_debug(GOSSIP_NAME_DEBUG, "%s called on %s\n",
__func__, dentry->d_name.name);
if (dentry->d_name.len > (ORANGEFS_NAME_MAX - 1))
return ERR_PTR(-ENAMETOOLONG);
new_op = op_alloc(ORANGEFS_VFS_OP_LOOKUP);
if (!new_op)
return ERR_PTR(-ENOMEM);
new_op->upcall.req.lookup.sym_follow = ORANGEFS_LOOKUP_LINK_NO_FOLLOW;
gossip_debug(GOSSIP_NAME_DEBUG, "%s:%s:%d using parent %pU\n",
__FILE__,
__func__,
__LINE__,
&parent->refn.khandle);
new_op->upcall.req.lookup.parent_refn = parent->refn;
strncpy(new_op->upcall.req.lookup.d_name, dentry->d_name.name,
ORANGEFS_NAME_MAX);
gossip_debug(GOSSIP_NAME_DEBUG,
"%s: doing lookup on %s under %pU,%d\n",
__func__,
new_op->upcall.req.lookup.d_name,
&new_op->upcall.req.lookup.parent_refn.khandle,
new_op->upcall.req.lookup.parent_refn.fs_id);
ret = service_operation(new_op, __func__, get_interruptible_flag(dir));
gossip_debug(GOSSIP_NAME_DEBUG,
"Lookup Got %pU, fsid %d (ret=%d)\n",
&new_op->downcall.resp.lookup.refn.khandle,
new_op->downcall.resp.lookup.refn.fs_id,
ret);
if (ret < 0) {
if (ret == -ENOENT) {
/*
* if no inode was found, add a negative dentry to
* dcache anyway; if we don't, we don't hold expected
* lookup semantics and we most noticeably break
* during directory renames.
*
* however, if the operation failed or exited, do not
* add the dentry (e.g. in the case that a touch is
* issued on a file that already exists that was
* interrupted during this lookup -- no need to add
* another negative dentry for an existing file)
*/
gossip_debug(GOSSIP_NAME_DEBUG,
"orangefs_lookup: Adding *negative* dentry "
"%p for %s\n",
dentry,
dentry->d_name.name);
d_add(dentry, NULL);
res = NULL;
goto out;
}
/* must be a non-recoverable error */
res = ERR_PTR(ret);
goto out;
}
dentry->d_time = jiffies + dcache_timeout_msecs*HZ/1000;
inode = orangefs_iget(dir->i_sb, &new_op->downcall.resp.lookup.refn);
if (IS_ERR(inode)) {
gossip_debug(GOSSIP_NAME_DEBUG,
"error %ld from iget\n", PTR_ERR(inode));
res = ERR_CAST(inode);
goto out;
}
ORANGEFS_I(inode)->getattr_time = jiffies - 1;
gossip_debug(GOSSIP_NAME_DEBUG,
"%s:%s:%d "
"Found good inode [%lu] with count [%d]\n",
__FILE__,
__func__,
__LINE__,
inode->i_ino,
(int)atomic_read(&inode->i_count));
/* update dentry/inode pair into dcache */
res = d_splice_alias(inode, dentry);
gossip_debug(GOSSIP_NAME_DEBUG,
"Lookup success (inode ct = %d)\n",
(int)atomic_read(&inode->i_count));
out:
op_release(new_op);
return res;
}
/*
* Get a newly allocated inode to go with a negative dentry.
*/
static int orangefs_create(struct inode *dir,
struct dentry *dentry,
umode_t mode,
bool exclusive)
{
struct orangefs_inode_s *parent = ORANGEFS_I(dir);
struct orangefs_kernel_op_s *new_op;
struct inode *inode;
int ret;
gossip_debug(GOSSIP_NAME_DEBUG, "%s: %s\n",
__func__,
dentry->d_name.name);
new_op = op_alloc(ORANGEFS_VFS_OP_CREATE);
if (!new_op)
return -ENOMEM;
new_op->upcall.req.create.parent_refn = parent->refn;
fill_default_sys_attrs(new_op->upcall.req.create.attributes,
ORANGEFS_TYPE_METAFILE, mode);
strncpy(new_op->upcall.req.create.d_name,
dentry->d_name.name, ORANGEFS_NAME_MAX);
ret = service_operation(new_op, __func__, get_interruptible_flag(dir));
gossip_debug(GOSSIP_NAME_DEBUG,
"%s: %s: handle:%pU: fsid:%d: new_op:%p: ret:%d:\n",
__func__,
dentry->d_name.name,
&new_op->downcall.resp.create.refn.khandle,
new_op->downcall.resp.create.refn.fs_id,
new_op,
ret);
if (ret < 0)
goto out;
inode = orangefs_new_inode(dir->i_sb, dir, S_IFREG | mode, 0,
&new_op->downcall.resp.create.refn);
if (IS_ERR(inode)) {
gossip_err("%s: Failed to allocate inode for file :%s:\n",
__func__,
dentry->d_name.name);
ret = PTR_ERR(inode);
goto out;
}
gossip_debug(GOSSIP_NAME_DEBUG,
"%s: Assigned inode :%pU: for file :%s:\n",
__func__,
get_khandle_from_ino(inode),
dentry->d_name.name);
d_instantiate(dentry, inode);
unlock_new_inode(inode);
dentry->d_time = jiffies + dcache_timeout_msecs*HZ/1000;
ORANGEFS_I(inode)->getattr_time = jiffies - 1;
gossip_debug(GOSSIP_NAME_DEBUG,
"%s: dentry instantiated for %s\n",
__func__,
dentry->d_name.name);
SetMtimeFlag(parent);
dir->i_mtime = dir->i_ctime = current_fs_time(dir->i_sb);
mark_inode_dirty_sync(dir);
ret = 0;
out:
op_release(new_op);
gossip_debug(GOSSIP_NAME_DEBUG,
"%s: %s: returning %d\n",
__func__,
dentry->d_name.name,
ret);
return ret;
}
/*
* Post and wait for the I/O upcall to finish
*/
static ssize_t wait_for_direct_io(enum ORANGEFS_io_type type, struct inode *inode,
loff_t *offset, struct iov_iter *iter,
size_t total_size, loff_t readahead_size)
{
struct orangefs_inode_s *orangefs_inode = ORANGEFS_I(inode);
struct orangefs_khandle *handle = &orangefs_inode->refn.khandle;
struct orangefs_kernel_op_s *new_op = NULL;
struct iov_iter saved = *iter;
int buffer_index = -1;
ssize_t ret;
new_op = op_alloc(ORANGEFS_VFS_OP_FILE_IO);
if (!new_op)
return -ENOMEM;
/* synchronous I/O */
new_op->upcall.req.io.readahead_size = readahead_size;
new_op->upcall.req.io.io_type = type;
new_op->upcall.req.io.refn = orangefs_inode->refn;
populate_shared_memory:
/* get a shared buffer index */
buffer_index = orangefs_bufmap_get();
if (buffer_index < 0) {
ret = buffer_index;
gossip_debug(GOSSIP_FILE_DEBUG,
"%s: orangefs_bufmap_get failure (%zd)\n",
__func__, ret);
goto out;
}
gossip_debug(GOSSIP_FILE_DEBUG,
"%s(%pU): GET op %p -> buffer_index %d\n",
__func__,
handle,
new_op,
buffer_index);
new_op->uses_shared_memory = 1;
new_op->upcall.req.io.buf_index = buffer_index;
new_op->upcall.req.io.count = total_size;
new_op->upcall.req.io.offset = *offset;
gossip_debug(GOSSIP_FILE_DEBUG,
"%s(%pU): offset: %llu total_size: %zd\n",
__func__,
handle,
llu(*offset),
total_size);
/*
* Stage 1: copy the buffers into client-core's address space
* precopy_buffers only pertains to writes.
*/
if (type == ORANGEFS_IO_WRITE) {
ret = precopy_buffers(buffer_index,
iter,
total_size);
if (ret < 0)
goto out;
}
gossip_debug(GOSSIP_FILE_DEBUG,
"%s(%pU): Calling post_io_request with tag (%llu)\n",
__func__,
handle,
llu(new_op->tag));
/* Stage 2: Service the I/O operation */
ret = service_operation(new_op,
type == ORANGEFS_IO_WRITE ?
"file_write" :
"file_read",
get_interruptible_flag(inode));
/*
* If service_operation() returns -EAGAIN #and# the operation was
* purged from orangefs_request_list or htable_ops_in_progress, then
* we know that the client was restarted, causing the shared memory
* area to be wiped clean. To restart a write operation in this
* case, we must re-copy the data from the user's iovec to a NEW
* shared memory location. To restart a read operation, we must get
* a new shared memory location.
*/
if (ret == -EAGAIN && op_state_purged(new_op)) {
orangefs_bufmap_put(buffer_index);
buffer_index = -1;
if (type == ORANGEFS_IO_WRITE)
*iter = saved;
gossip_debug(GOSSIP_FILE_DEBUG,
"%s:going to repopulate_shared_memory.\n",
__func__);
goto populate_shared_memory;
}
if (ret < 0) {
if (ret == -EINTR) {
/*
* We can't return EINTR if any data was written,
* it's not POSIX. It is minimally acceptable
* to give a partial write, the way NFS does.
*
* It would be optimal to return all or nothing,
* but if a userspace write is bigger than
* an IO buffer, and the interrupt occurs
* between buffer writes, that would not be
* possible.
*/
switch (new_op->op_state - OP_VFS_STATE_GIVEN_UP) {
/*
* If the op was waiting when the interrupt
* occurred, then the client-core did not
* trigger the write.
*/
case OP_VFS_STATE_WAITING:
if (*offset == 0)
ret = -EINTR;
else
ret = 0;
break;
/*
* If the op was in progress when the interrupt
* occurred, then the client-core was able to
* trigger the write.
*/
case OP_VFS_STATE_INPROGR:
ret = total_size;
break;
default:
gossip_err("%s: unexpected op state :%d:.\n",
__func__,
new_op->op_state);
ret = 0;
break;
}
gossip_debug(GOSSIP_FILE_DEBUG,
"%s: got EINTR, state:%d: %p\n",
__func__,
new_op->op_state,
new_op);
} else {
gossip_err("%s: error in %s handle %pU, returning %zd\n",
__func__,
type == ORANGEFS_IO_READ ?
"read from" : "write to",
handle, ret);
}
if (orangefs_cancel_op_in_progress(new_op))
return ret;
goto out;
}
/*
* Stage 3: Post copy buffers from client-core's address space
* postcopy_buffers only pertains to reads.
*/
if (type == ORANGEFS_IO_READ) {
ret = postcopy_buffers(buffer_index,
iter,
new_op->downcall.resp.io.amt_complete);
if (ret < 0)
goto out;
}
gossip_debug(GOSSIP_FILE_DEBUG,
"%s(%pU): Amount %s, returned by the sys-io call:%d\n",
__func__,
handle,
type == ORANGEFS_IO_READ ? "read" : "written",
(int)new_op->downcall.resp.io.amt_complete);
ret = new_op->downcall.resp.io.amt_complete;
out:
if (buffer_index >= 0) {
orangefs_bufmap_put(buffer_index);
gossip_debug(GOSSIP_FILE_DEBUG,
"%s(%pU): PUT buffer_index %d\n",
__func__, handle, buffer_index);
buffer_index = -1;
}
op_release(new_op);
return ret;
}
struct dentry *orangefs_mount(struct file_system_type *fst,
int flags,
const char *devname,
void *data)
{
int ret = -EINVAL;
struct super_block *sb = ERR_PTR(-EINVAL);
struct orangefs_kernel_op_s *new_op;
struct dentry *d = ERR_PTR(-EINVAL);
gossip_debug(GOSSIP_SUPER_DEBUG,
"orangefs_mount: called with devname %s\n",
devname);
if (!devname) {
gossip_err("ERROR: device name not specified.\n");
return ERR_PTR(-EINVAL);
}
new_op = op_alloc(ORANGEFS_VFS_OP_FS_MOUNT);
if (!new_op)
return ERR_PTR(-ENOMEM);
strncpy(new_op->upcall.req.fs_mount.orangefs_config_server,
devname,
ORANGEFS_MAX_SERVER_ADDR_LEN - 1);
gossip_debug(GOSSIP_SUPER_DEBUG,
"Attempting ORANGEFS Mount via host %s\n",
new_op->upcall.req.fs_mount.orangefs_config_server);
ret = service_operation(new_op, "orangefs_mount", 0);
gossip_debug(GOSSIP_SUPER_DEBUG,
"orangefs_mount: mount got return value of %d\n", ret);
if (ret)
goto free_op;
if (new_op->downcall.resp.fs_mount.fs_id == ORANGEFS_FS_ID_NULL) {
gossip_err("ERROR: Retrieved null fs_id\n");
ret = -EINVAL;
goto free_op;
}
sb = sget(fst, NULL, set_anon_super, flags, NULL);
if (IS_ERR(sb)) {
d = ERR_CAST(sb);
orangefs_unmount(new_op->downcall.resp.fs_mount.id,
new_op->downcall.resp.fs_mount.fs_id, devname);
goto free_op;
}
ret = orangefs_fill_sb(sb,
&new_op->downcall.resp.fs_mount, data,
flags & SB_SILENT ? 1 : 0);
if (ret) {
d = ERR_PTR(ret);
goto free_sb_and_op;
}
/*
* on successful mount, store the devname and data
* used
*/
strncpy(ORANGEFS_SB(sb)->devname,
devname,
ORANGEFS_MAX_SERVER_ADDR_LEN - 1);
/* mount_pending must be cleared */
ORANGEFS_SB(sb)->mount_pending = 0;
/*
* finally, add this sb to our list of known orangefs
* sb's
*/
gossip_debug(GOSSIP_SUPER_DEBUG,
"Adding SB %p to orangefs superblocks\n",
ORANGEFS_SB(sb));
spin_lock(&orangefs_superblocks_lock);
list_add_tail(&ORANGEFS_SB(sb)->list, &orangefs_superblocks);
spin_unlock(&orangefs_superblocks_lock);
op_release(new_op);
/* Must be removed from the list now. */
ORANGEFS_SB(sb)->no_list = 0;
if (orangefs_userspace_version >= 20906) {
new_op = op_alloc(ORANGEFS_VFS_OP_FEATURES);
if (!new_op)
return ERR_PTR(-ENOMEM);
new_op->upcall.req.features.features = 0;
ret = service_operation(new_op, "orangefs_features", 0);
orangefs_features = new_op->downcall.resp.features.features;
op_release(new_op);
} else {
orangefs_features = 0;
}
return dget(sb->s_root);
free_sb_and_op:
/* Will call orangefs_kill_sb with sb not in list. */
ORANGEFS_SB(sb)->no_list = 1;
/* ORANGEFS_VFS_OP_FS_UMOUNT is done by orangefs_kill_sb. */
deactivate_locked_super(sb);
free_op:
gossip_err("orangefs_mount: mount request failed with %d\n", ret);
if (ret == -EINVAL) {
gossip_err("Ensure that all orangefs-servers have the same FS configuration files\n");
gossip_err("Look at pvfs2-client-core log file (typically /tmp/pvfs2-client.log) for more details\n");
}
op_release(new_op);
return d;
}
static int orangefs_mkdir(struct inode *dir, struct dentry *dentry, umode_t mode)
{
struct orangefs_inode_s *parent = ORANGEFS_I(dir);
struct orangefs_kernel_op_s *new_op;
struct inode *inode;
int ret;
new_op = op_alloc(ORANGEFS_VFS_OP_MKDIR);
if (!new_op)
return -ENOMEM;
new_op->upcall.req.mkdir.parent_refn = parent->refn;
fill_default_sys_attrs(new_op->upcall.req.mkdir.attributes,
ORANGEFS_TYPE_DIRECTORY, mode);
strncpy(new_op->upcall.req.mkdir.d_name,
dentry->d_name.name, ORANGEFS_NAME_MAX);
ret = service_operation(new_op, __func__, get_interruptible_flag(dir));
gossip_debug(GOSSIP_NAME_DEBUG,
"Mkdir Got ORANGEFS handle %pU on fsid %d\n",
&new_op->downcall.resp.mkdir.refn.khandle,
new_op->downcall.resp.mkdir.refn.fs_id);
if (ret < 0) {
gossip_debug(GOSSIP_NAME_DEBUG,
"%s: failed with error code %d\n",
__func__, ret);
goto out;
}
inode = orangefs_new_inode(dir->i_sb, dir, S_IFDIR | mode, 0,
&new_op->downcall.resp.mkdir.refn);
if (IS_ERR(inode)) {
gossip_err("*** Failed to allocate orangefs dir inode\n");
ret = PTR_ERR(inode);
goto out;
}
gossip_debug(GOSSIP_NAME_DEBUG,
"Assigned dir inode new number of %pU\n",
get_khandle_from_ino(inode));
d_instantiate(dentry, inode);
unlock_new_inode(inode);
dentry->d_time = jiffies + dcache_timeout_msecs*HZ/1000;
ORANGEFS_I(inode)->getattr_time = jiffies - 1;
gossip_debug(GOSSIP_NAME_DEBUG,
"Inode (Directory) %pU -> %s\n",
get_khandle_from_ino(inode),
dentry->d_name.name);
/*
* NOTE: we have no good way to keep nlink consistent for directories
* across clients; keep constant at 1.
*/
SetMtimeFlag(parent);
dir->i_mtime = dir->i_ctime = current_fs_time(dir->i_sb);
mark_inode_dirty_sync(dir);
out:
op_release(new_op);
return ret;
}
/*
* Tries to get a specified key's attributes of a given
* file into a user-specified buffer. Note that the getxattr
* interface allows for the users to probe the size of an
* extended attribute by passing in a value of 0 to size.
* Thus our return value is always the size of the attribute
* unless the key does not exist for the file and/or if
* there were errors in fetching the attribute value.
*/
ssize_t orangefs_inode_getxattr(struct inode *inode, const char *name,
void *buffer, size_t size)
{
struct orangefs_inode_s *orangefs_inode = ORANGEFS_I(inode);
struct orangefs_kernel_op_s *new_op = NULL;
ssize_t ret = -ENOMEM;
ssize_t length = 0;
int fsuid;
int fsgid;
gossip_debug(GOSSIP_XATTR_DEBUG,
"%s: name %s, buffer_size %zd\n",
__func__, name, size);
if (S_ISLNK(inode->i_mode))
return -EOPNOTSUPP;
if (strlen(name) >= ORANGEFS_MAX_XATTR_NAMELEN)
return -EINVAL;
fsuid = from_kuid(&init_user_ns, current_fsuid());
fsgid = from_kgid(&init_user_ns, current_fsgid());
gossip_debug(GOSSIP_XATTR_DEBUG,
"getxattr on inode %pU, name %s "
"(uid %o, gid %o)\n",
get_khandle_from_ino(inode),
name,
fsuid,
fsgid);
down_read(&orangefs_inode->xattr_sem);
new_op = op_alloc(ORANGEFS_VFS_OP_GETXATTR);
if (!new_op)
goto out_unlock;
new_op->upcall.req.getxattr.refn = orangefs_inode->refn;
strcpy(new_op->upcall.req.getxattr.key, name);
/*
* NOTE: Although keys are meant to be NULL terminated textual
* strings, I am going to explicitly pass the length just in case
* we change this later on...
*/
new_op->upcall.req.getxattr.key_sz = strlen(name) + 1;
ret = service_operation(new_op, "orangefs_inode_getxattr",
get_interruptible_flag(inode));
if (ret != 0) {
if (ret == -ENOENT) {
ret = -ENODATA;
gossip_debug(GOSSIP_XATTR_DEBUG,
"orangefs_inode_getxattr: inode %pU key %s"
" does not exist!\n",
get_khandle_from_ino(inode),
(char *)new_op->upcall.req.getxattr.key);
}
goto out_release_op;
}
/*
* Length returned includes null terminator.
*/
length = new_op->downcall.resp.getxattr.val_sz;
/*
* Just return the length of the queried attribute.
*/
if (size == 0) {
ret = length;
goto out_release_op;
}
/*
* Check to see if key length is > provided buffer size.
*/
if (length > size) {
ret = -ERANGE;
goto out_release_op;
}
memcpy(buffer, new_op->downcall.resp.getxattr.val, length);
memset(buffer + length, 0, size - length);
gossip_debug(GOSSIP_XATTR_DEBUG,
"orangefs_inode_getxattr: inode %pU "
"key %s key_sz %d, val_len %d\n",
get_khandle_from_ino(inode),
(char *)new_op->
upcall.req.getxattr.key,
(int)new_op->
upcall.req.getxattr.key_sz,
(int)ret);
ret = length;
out_release_op:
op_release(new_op);
out_unlock:
up_read(&orangefs_inode->xattr_sem);
return ret;
}
static struct dentry *pvfs2_lookup(
struct inode *dir,
struct dentry *dentry,
struct nameidata *nd)
#endif
{
int ret = -EINVAL;
struct inode *inode = NULL;
pvfs2_kernel_op_t *new_op = NULL;
pvfs2_inode_t *parent = NULL, *found_pvfs2_inode = NULL;
struct super_block *sb = NULL;
/*
in theory we could skip a lookup here (if the intent is to
create) in order to avoid a potentially failed lookup, but
leaving it in can skip a valid lookup and try to create a file
that already exists (e.g. the vfs already handles checking for
-EEXIST on O_EXCL opens, which is broken if we skip this lookup
in the create path)
*/
gossip_debug(GOSSIP_NAME_DEBUG, "pvfs2_lookup called on %s\n", dentry->d_name.name);
if (dentry->d_name.len > (PVFS2_NAME_LEN-1))
{
return ERR_PTR(-ENAMETOOLONG);
}
new_op = op_alloc(PVFS2_VFS_OP_LOOKUP);
if (!new_op)
{
return ERR_PTR(-ENOMEM);
}
#ifdef PVFS2_LINUX_KERNEL_2_4
new_op->upcall.req.lookup.sym_follow = PVFS2_LOOKUP_LINK_NO_FOLLOW;
#else
/*
if we're at a symlink, should we follow it? never attempt to
follow negative dentries
*/
new_op->upcall.req.lookup.sym_follow =
((nd && (nd->flags & LOOKUP_FOLLOW) &&
(dentry->d_inode != NULL)) ?
PVFS2_LOOKUP_LINK_FOLLOW : PVFS2_LOOKUP_LINK_NO_FOLLOW);
#endif
if (dir)
{
sb = dir->i_sb;
parent = PVFS2_I(dir);
if (parent && parent->refn.handle != PVFS_HANDLE_NULL
&& parent->refn.fs_id != PVFS_FS_ID_NULL)
{
gossip_debug(GOSSIP_NAME_DEBUG, "%s:%s:%d using parent %llu\n",
__FILE__, __func__, __LINE__, llu(parent->refn.handle));
new_op->upcall.req.lookup.parent_refn = parent->refn;
}
else
{
#if defined(HAVE_IGET4_LOCKED) || defined(HAVE_IGET5_LOCKED)
gossip_lerr("Critical error: i_ino cannot be relied on when using iget5/iget4\n");
op_release(new_op);
return ERR_PTR(-EINVAL);
#endif
new_op->upcall.req.lookup.parent_refn.handle =
get_handle_from_ino(dir);
new_op->upcall.req.lookup.parent_refn.fs_id =
PVFS2_SB(sb)->fs_id;
}
}
else
{
/*
if no parent at all was provided, use the root
handle and file system id stored in the super
block for the specified dentry's inode
*/
sb = dentry->d_inode->i_sb;
new_op->upcall.req.lookup.parent_refn.handle =
PVFS2_SB(sb)->root_handle;
new_op->upcall.req.lookup.parent_refn.fs_id =
PVFS2_SB(sb)->fs_id;
}
strncpy(new_op->upcall.req.lookup.d_name,
dentry->d_name.name, PVFS2_NAME_LEN);
gossip_debug(GOSSIP_NAME_DEBUG, "pvfs2_lookup: doing lookup on %s\n under %llu,%d "
"(follow=%s)\n", new_op->upcall.req.lookup.d_name,
llu(new_op->upcall.req.lookup.parent_refn.handle),
new_op->upcall.req.lookup.parent_refn.fs_id,
((new_op->upcall.req.lookup.sym_follow ==
PVFS2_LOOKUP_LINK_FOLLOW) ? "yes" : "no"));
ret = service_operation(
new_op, "pvfs2_lookup",
get_interruptible_flag(dir));
gossip_debug(GOSSIP_NAME_DEBUG, "Lookup Got %llu, fsid %d (ret=%d)\n",
llu(new_op->downcall.resp.lookup.refn.handle),
new_op->downcall.resp.lookup.refn.fs_id, ret);
if(ret < 0)
{
if(ret == -ENOENT)
{
/*
* if no inode was found, add a negative dentry to dcache anyway;
* if we don't, we don't hold expected lookup semantics and we most
* noticeably break during directory renames.
*
* however, if the operation failed or exited, do not add the
* dentry (e.g. in the case that a touch is issued on a file that
* already exists that was interrupted during this lookup -- no
* need to add another negative dentry for an existing file)
*/
gossip_debug(GOSSIP_NAME_DEBUG,
"pvfs2_lookup: Adding *negative* dentry %p\n for %s\n",
dentry, dentry->d_name.name);
/*
* make sure to set the pvfs2 specific dentry operations for
* the negative dentry that we're adding now so that a
* potential future lookup of this cached negative dentry can
* be properly revalidated.
*/
dentry->d_op = &pvfs2_dentry_operations;
d_add(dentry, inode);
op_release(new_op);
return NULL;
}
op_release(new_op);
/* must be a non-recoverable error */
return ERR_PTR(ret);
}
inode = pvfs2_iget(sb, &new_op->downcall.resp.lookup.refn);
if (inode && !is_bad_inode(inode))
{
struct dentry *res;
gossip_debug(GOSSIP_NAME_DEBUG, "%s:%s:%d Found good inode [%lu] with count [%d]\n",
__FILE__, __func__, __LINE__, inode->i_ino, (int)atomic_read(&inode->i_count));
/* update dentry/inode pair into dcache */
dentry->d_op = &pvfs2_dentry_operations;
res = pvfs2_d_splice_alias(dentry, inode);
gossip_debug(GOSSIP_NAME_DEBUG, "Lookup success (inode ct = %d)\n",
(int)atomic_read(&inode->i_count));
if (res)
res->d_op = &pvfs2_dentry_operations;
op_release(new_op);
#ifdef PVFS2_LINUX_KERNEL_2_4
return NULL;
#else
return res;
#endif
}
else if (inode && is_bad_inode(inode))
{
gossip_debug(GOSSIP_NAME_DEBUG, "%s:%s:%d Found bad inode [%lu] with count [%d]. Returning error [%d]",
__FILE__, __func__, __LINE__, inode->i_ino, (int)atomic_read(&inode->i_count), ret);
ret = -EACCES;
found_pvfs2_inode = PVFS2_I(inode);
/* look for an error code, possibly set by pvfs2_read_inode(),
* otherwise we have to guess EACCES
*/
if(found_pvfs2_inode->error_code)
{
ret = found_pvfs2_inode->error_code;
}
iput(inode);
op_release(new_op);
return ERR_PTR(ret);
}
/* no error was returned from service_operation, but the inode
* from pvfs2_iget was null...just return EACCESS
*/
op_release(new_op);
gossip_debug(GOSSIP_NAME_DEBUG, "Returning -EACCES for NULL inode\n");
return ERR_PTR(-EACCES);
}
static int orangefs_symlink(struct inode *dir,
struct dentry *dentry,
const char *symname)
{
struct orangefs_inode_s *parent = ORANGEFS_I(dir);
struct orangefs_kernel_op_s *new_op;
struct inode *inode;
int mode = 755;
int ret;
gossip_debug(GOSSIP_NAME_DEBUG, "%s: called\n", __func__);
if (!symname)
return -EINVAL;
if (strlen(symname)+1 > ORANGEFS_NAME_MAX)
return -ENAMETOOLONG;
new_op = op_alloc(ORANGEFS_VFS_OP_SYMLINK);
if (!new_op)
return -ENOMEM;
new_op->upcall.req.sym.parent_refn = parent->refn;
fill_default_sys_attrs(new_op->upcall.req.sym.attributes,
ORANGEFS_TYPE_SYMLINK,
mode);
strncpy(new_op->upcall.req.sym.entry_name,
dentry->d_name.name,
ORANGEFS_NAME_MAX);
strncpy(new_op->upcall.req.sym.target, symname, ORANGEFS_NAME_MAX);
ret = service_operation(new_op, __func__, get_interruptible_flag(dir));
gossip_debug(GOSSIP_NAME_DEBUG,
"Symlink Got ORANGEFS handle %pU on fsid %d (ret=%d)\n",
&new_op->downcall.resp.sym.refn.khandle,
new_op->downcall.resp.sym.refn.fs_id, ret);
if (ret < 0) {
gossip_debug(GOSSIP_NAME_DEBUG,
"%s: failed with error code %d\n",
__func__, ret);
goto out;
}
inode = orangefs_new_inode(dir->i_sb, dir, S_IFLNK | mode, 0,
&new_op->downcall.resp.sym.refn);
if (IS_ERR(inode)) {
gossip_err
("*** Failed to allocate orangefs symlink inode\n");
ret = PTR_ERR(inode);
goto out;
}
gossip_debug(GOSSIP_NAME_DEBUG,
"Assigned symlink inode new number of %pU\n",
get_khandle_from_ino(inode));
d_instantiate(dentry, inode);
unlock_new_inode(inode);
dentry->d_time = jiffies + dcache_timeout_msecs*HZ/1000;
ORANGEFS_I(inode)->getattr_time = jiffies - 1;
gossip_debug(GOSSIP_NAME_DEBUG,
"Inode (Symlink) %pU -> %s\n",
get_khandle_from_ino(inode),
dentry->d_name.name);
SetMtimeFlag(parent);
dir->i_mtime = dir->i_ctime = current_fs_time(dir->i_sb);
mark_inode_dirty_sync(dir);
ret = 0;
out:
op_release(new_op);
return ret;
}
/*
* Tries to set an attribute for a given key on a file.
*
* Returns a -ve number on error and 0 on success. Key is text, but value
* can be binary!
*/
int orangefs_inode_setxattr(struct inode *inode, const char *name,
const void *value, size_t size, int flags)
{
struct orangefs_inode_s *orangefs_inode = ORANGEFS_I(inode);
struct orangefs_kernel_op_s *new_op;
int internal_flag = 0;
int ret = -ENOMEM;
gossip_debug(GOSSIP_XATTR_DEBUG,
"%s: name %s, buffer_size %zd\n",
__func__, name, size);
if (size > ORANGEFS_MAX_XATTR_VALUELEN)
return -EINVAL;
if (strlen(name) >= ORANGEFS_MAX_XATTR_NAMELEN)
return -EINVAL;
internal_flag = convert_to_internal_xattr_flags(flags);
/* This is equivalent to a removexattr */
if (size == 0 && !value) {
gossip_debug(GOSSIP_XATTR_DEBUG,
"removing xattr (%s)\n",
name);
return orangefs_inode_removexattr(inode, name, flags);
}
gossip_debug(GOSSIP_XATTR_DEBUG,
"setxattr on inode %pU, name %s\n",
get_khandle_from_ino(inode),
name);
down_write(&orangefs_inode->xattr_sem);
new_op = op_alloc(ORANGEFS_VFS_OP_SETXATTR);
if (!new_op)
goto out_unlock;
new_op->upcall.req.setxattr.refn = orangefs_inode->refn;
new_op->upcall.req.setxattr.flags = internal_flag;
/*
* NOTE: Although keys are meant to be NULL terminated textual
* strings, I am going to explicitly pass the length just in
* case we change this later on...
*/
strcpy(new_op->upcall.req.setxattr.keyval.key, name);
new_op->upcall.req.setxattr.keyval.key_sz = strlen(name) + 1;
memcpy(new_op->upcall.req.setxattr.keyval.val, value, size);
new_op->upcall.req.setxattr.keyval.val_sz = size;
gossip_debug(GOSSIP_XATTR_DEBUG,
"orangefs_inode_setxattr: key %s, key_sz %d "
" value size %zd\n",
(char *)new_op->upcall.req.setxattr.keyval.key,
(int)new_op->upcall.req.setxattr.keyval.key_sz,
size);
ret = service_operation(new_op,
"orangefs_inode_setxattr",
get_interruptible_flag(inode));
gossip_debug(GOSSIP_XATTR_DEBUG,
"orangefs_inode_setxattr: returning %d\n",
ret);
/* when request is serviced properly, free req op struct */
op_release(new_op);
out_unlock:
up_write(&orangefs_inode->xattr_sem);
return ret;
}
/*
* Tries to get a specified object's keys into a user-specified buffer of a
* given size. Note that like the previous instances of xattr routines, this
* also allows you to pass in a NULL pointer and 0 size to probe the size for
* subsequent memory allocations. Thus our return value is always the size of
* all the keys unless there were errors in fetching the keys!
*/
ssize_t orangefs_listxattr(struct dentry *dentry, char *buffer, size_t size)
{
struct inode *inode = dentry->d_inode;
struct orangefs_inode_s *orangefs_inode = ORANGEFS_I(inode);
struct orangefs_kernel_op_s *new_op;
__u64 token = ORANGEFS_ITERATE_START;
ssize_t ret = -ENOMEM;
ssize_t total = 0;
int count_keys = 0;
int key_size;
int i = 0;
int returned_count = 0;
if (size > 0 && !buffer) {
gossip_err("%s: bogus NULL pointers\n", __func__);
return -EINVAL;
}
down_read(&orangefs_inode->xattr_sem);
new_op = op_alloc(ORANGEFS_VFS_OP_LISTXATTR);
if (!new_op)
goto out_unlock;
if (buffer && size > 0)
memset(buffer, 0, size);
try_again:
key_size = 0;
new_op->upcall.req.listxattr.refn = orangefs_inode->refn;
new_op->upcall.req.listxattr.token = token;
new_op->upcall.req.listxattr.requested_count =
(size == 0) ? 0 : ORANGEFS_MAX_XATTR_LISTLEN;
ret = service_operation(new_op, __func__,
get_interruptible_flag(inode));
if (ret != 0)
goto done;
if (size == 0) {
/*
* This is a bit of a big upper limit, but I did not want to
* spend too much time getting this correct, since users end
* up allocating memory rather than us...
*/
total = new_op->downcall.resp.listxattr.returned_count *
ORANGEFS_MAX_XATTR_NAMELEN;
goto done;
}
returned_count = new_op->downcall.resp.listxattr.returned_count;
if (returned_count < 0 ||
returned_count > ORANGEFS_MAX_XATTR_LISTLEN) {
gossip_err("%s: impossible value for returned_count:%d:\n",
__func__,
returned_count);
ret = -EIO;
goto done;
}
/*
* Check to see how much can be fit in the buffer. Fit only whole keys.
*/
for (i = 0; i < returned_count; i++) {
if (new_op->downcall.resp.listxattr.lengths[i] < 0 ||
new_op->downcall.resp.listxattr.lengths[i] >
ORANGEFS_MAX_XATTR_NAMELEN) {
gossip_err("%s: impossible value for lengths[%d]\n",
__func__,
new_op->downcall.resp.listxattr.lengths[i]);
ret = -EIO;
goto done;
}
if (total + new_op->downcall.resp.listxattr.lengths[i] > size)
goto done;
/*
* Since many dumb programs try to setxattr() on our reserved
* xattrs this is a feeble attempt at defeating those by not
* listing them in the output of listxattr.. sigh
*/
if (is_reserved_key(new_op->downcall.resp.listxattr.key +
key_size,
new_op->downcall.resp.
listxattr.lengths[i])) {
gossip_debug(GOSSIP_XATTR_DEBUG, "Copying key %d -> %s\n",
i, new_op->downcall.resp.listxattr.key +
key_size);
memcpy(buffer + total,
new_op->downcall.resp.listxattr.key + key_size,
new_op->downcall.resp.listxattr.lengths[i]);
total += new_op->downcall.resp.listxattr.lengths[i];
count_keys++;
} else {
gossip_debug(GOSSIP_XATTR_DEBUG, "[RESERVED] key %d -> %s\n",
i, new_op->downcall.resp.listxattr.key +
key_size);
}
key_size += new_op->downcall.resp.listxattr.lengths[i];
}
/*
* Since the buffer was large enough, we might have to continue
* fetching more keys!
*/
token = new_op->downcall.resp.listxattr.token;
if (token != ORANGEFS_ITERATE_END)
goto try_again;
done:
gossip_debug(GOSSIP_XATTR_DEBUG, "%s: returning %d"
" [size of buffer %ld] (filled in %d keys)\n",
__func__,
ret ? (int)ret : (int)total,
(long)size,
count_keys);
op_release(new_op);
if (ret == 0)
ret = total;
out_unlock:
up_read(&orangefs_inode->xattr_sem);
return ret;
}
static int pvfs2_rename(
struct inode *old_dir,
struct dentry *old_dentry,
struct inode *new_dir,
struct dentry *new_dentry)
{
int ret = -EINVAL, are_directories = 0;
pvfs2_inode_t *pvfs2_old_parent_inode = PVFS2_I(old_dir);
pvfs2_inode_t *pvfs2_new_parent_inode = PVFS2_I(new_dir);
pvfs2_kernel_op_t *new_op = NULL;
struct super_block *sb = NULL;
gossip_debug(GOSSIP_NAME_DEBUG, "pvfs2_rename: called (%s/%s => %s/%s) ct=%d\n",
old_dentry->d_parent->d_name.name, old_dentry->d_name.name,
new_dentry->d_parent->d_name.name, new_dentry->d_name.name,
atomic_read(&new_dentry->d_count));
are_directories = S_ISDIR(old_dentry->d_inode->i_mode);
#if 0
/* NOTE: we have no good way to keep nlink consistent for directories
* across clients; keep constant at 1 -Phil
*/
if (are_directories && (new_dir->i_nlink >= PVFS2_LINK_MAX))
{
gossip_err("pvfs2_rename: directory %s surpassed "
"PVFS2_LINK_MAX\n",
new_dentry->d_name.name);
return -EMLINK;
}
#endif
new_op = op_alloc(PVFS2_VFS_OP_RENAME);
if (!new_op)
{
return ret;
}
/*
if no handle/fs_id is available in the parent,
use the root handle/fs_id as specified by the
inode's corresponding superblock
*/
if (pvfs2_old_parent_inode &&
pvfs2_old_parent_inode->refn.handle != PVFS_HANDLE_NULL &&
pvfs2_old_parent_inode->refn.fs_id != PVFS_FS_ID_NULL)
{
new_op->upcall.req.rename.old_parent_refn =
pvfs2_old_parent_inode->refn;
}
else
{
sb = old_dir->i_sb;
new_op->upcall.req.rename.old_parent_refn.handle =
PVFS2_SB(sb)->root_handle;
new_op->upcall.req.rename.old_parent_refn.fs_id =
PVFS2_SB(sb)->fs_id;
}
/* do the same for the new parent */
if (pvfs2_new_parent_inode &&
pvfs2_new_parent_inode->refn.handle != PVFS_HANDLE_NULL &&
pvfs2_new_parent_inode->refn.fs_id != PVFS_FS_ID_NULL)
{
new_op->upcall.req.rename.new_parent_refn =
pvfs2_new_parent_inode->refn;
}
else
{
sb = new_dir->i_sb;
new_op->upcall.req.rename.new_parent_refn.handle =
PVFS2_SB(sb)->root_handle;
new_op->upcall.req.rename.new_parent_refn.fs_id =
PVFS2_SB(sb)->fs_id;
}
strncpy(new_op->upcall.req.rename.d_old_name,
old_dentry->d_name.name, PVFS2_NAME_LEN);
strncpy(new_op->upcall.req.rename.d_new_name,
new_dentry->d_name.name, PVFS2_NAME_LEN);
ret = service_operation(
new_op, "pvfs2_rename",
get_interruptible_flag(old_dentry->d_inode));
gossip_debug(GOSSIP_NAME_DEBUG, "pvfs2_rename: got downcall status %d\n", ret);
if (new_dentry->d_inode)
{
new_dentry->d_inode->i_ctime = CURRENT_TIME;
#if 0
/* NOTE: we have no good way to keep nlink consistent for directories
* across clients; keep constant at 1 -Phil
*/
if (are_directories)
{
new_dentry->d_inode->i_nlink--;
}
#endif
}
#if 0
/* NOTE: we have no good way to keep nlink consistent for directories
* across clients; keep constant at 1 -Phil
*/
else if (are_directories)
{
new_dir->i_nlink++;
old_dir->i_nlink--;
}
#endif
op_release(new_op);
#ifdef PVFS2_LINUX_KERNEL_2_4
if (ret == 0)
{
d_move(old_dentry, new_dentry);
}
#endif
return ret;
}
