#include <linux/fs.h>
#include <linux/sched.h>
#include <linux/mm.h>
+#include <linux/file.h>
#include <linux/quotaops.h>
#include <linux/highmem.h>
#include <linux/pagemap.h>
-#include <linux/security.h>
#include <linux/user_namespace.h>
#include <linux/uidgid.h>
continue;
spin_lock_nested(&child->d_lock, DENTRY_D_LOCK_NESTED);
- __d_lustre_invalidate(child);
+ if (lld_is_init(child))
+ ll_d2d(child)->lld_invalid = 1;
if (!ll_d_count(child)) {
dget_dlock(child);
__d_drop(child);
}
if (bits & MDS_INODELOCK_XATTR) {
- ll_xattr_cache_destroy(inode);
+ ll_xattr_cache_empty(inode);
bits &= ~MDS_INODELOCK_XATTR;
}
struct dentry *ll_splice_alias(struct inode *inode, struct dentry *de)
{
struct dentry *new;
- int rc;
if (inode) {
new = ll_find_alias(inode, de);
if (new) {
- rc = ll_d_init(new);
- if (rc < 0) {
- dput(new);
- return ERR_PTR(rc);
- }
+ if (!ll_d_setup(new, true))
+ return ERR_PTR(-ENOMEM);
d_move(new, de);
iput(inode);
CDEBUG(D_DENTRY,
return new;
}
}
- rc = ll_d_init(de);
- if (rc < 0)
- return ERR_PTR(rc);
+ if (!ll_d_setup(de, false))
+ return ERR_PTR(-ENOMEM);
d_add(de, inode);
/* this needs only to be done for foreign symlink dirs as
struct mdt_body *body = req_capsule_server_get(pill,
&RMF_MDT_BODY);
- rc = ll_prep_inode(&inode, request, (*de)->d_sb, it);
+ rc = ll_prep_inode(&inode, &request->rq_pill, (*de)->d_sb, it);
if (rc)
RETURN(rc);
if (!it_disposition(it, DISP_LOOKUP_NEG)) {
/* we have lookup look - unhide dentry */
- if (bits & MDS_INODELOCK_LOOKUP)
+ if (bits & MDS_INODELOCK_LOOKUP) {
d_lustre_revalidate(*de);
+ ll_update_dir_depth(parent, (*de)->d_inode);
+ }
if (encrypt) {
rc = llcrypt_get_encryption_info(inode);
out:
if (rc != 0 && it->it_op & IT_OPEN) {
ll_intent_drop_lock(it);
- ll_open_cleanup((*de)->d_sb, request);
+ ll_open_cleanup((*de)->d_sb, &request->rq_pill);
}
return rc;
__u32 opc;
int rc;
char secctx_name[XATTR_NAME_MAX + 1];
-
+ struct llcrypt_name fname;
+ struct lu_fid fid;
ENTRY;
if (dentry->d_name.len > ll_i2sbi(parent)->ll_namelen)
if (it->it_op & IT_CREAT)
opc = LUSTRE_OPC_CREATE;
else
- opc = LUSTRE_OPC_ANY;
+ opc = LUSTRE_OPC_LOOKUP;
+
+ /* Here we should be calling llcrypt_prepare_lookup(). But it installs a
+ * custom ->d_revalidate() method, so we lose ll_d_ops.
+ * To workaround this, call ll_setup_filename() and do the rest
+ * manually. Also make a copy of llcrypt_d_revalidate() (unfortunately
+ * not exported function) and call it from ll_revalidate_dentry(), to
+ * ensure we do not cache stale dentries after a key has been added.
+ */
+ rc = ll_setup_filename(parent, &dentry->d_name, 1, &fname, &fid);
+ if ((!rc || rc == -ENOENT) && fname.is_ciphertext_name) {
+ spin_lock(&dentry->d_lock);
+ dentry->d_flags |= DCACHE_ENCRYPTED_NAME;
+ spin_unlock(&dentry->d_lock);
+ }
+ if (rc == -ENOENT)
+ RETURN(NULL);
+ if (rc)
+ RETURN(ERR_PTR(rc));
- op_data = ll_prep_md_op_data(NULL, parent, NULL, dentry->d_name.name,
- dentry->d_name.len, 0, opc, NULL);
- if (IS_ERR(op_data))
- GOTO(out, retval = ERR_CAST(op_data));
+ op_data = ll_prep_md_op_data(NULL, parent, NULL, fname.disk_name.name,
+ fname.disk_name.len, 0, opc, NULL);
+ if (IS_ERR(op_data)) {
+ llcrypt_free_filename(&fname);
+ RETURN(ERR_CAST(op_data));
+ }
+ if (!fid_is_zero(&fid)) {
+ op_data->op_fid2 = fid;
+ op_data->op_bias = MDS_FID_OP;
+ if (it->it_op & IT_OPEN)
+ it->it_flags |= MDS_OPEN_BY_FID;
+ }
/* enforce umask if acl disabled or MDS doesn't support umask */
if (!IS_POSIXACL(parent) || !exp_connect_umask(ll_i2mdexp(parent)))
it->it_create_mode &= ~current_umask();
if (it->it_op & IT_CREAT &&
- ll_i2sbi(parent)->ll_flags & LL_SBI_FILE_SECCTX) {
+ test_bit(LL_SBI_FILE_SECCTX, ll_i2sbi(parent)->ll_flags)) {
rc = ll_dentry_init_security(dentry, it->it_create_mode,
&dentry->d_name,
&op_data->op_file_secctx_name,
*secctxlen = 0;
}
if (it->it_op & IT_CREAT && encrypt) {
- rc = llcrypt_inherit_context(parent, NULL, op_data, false);
- if (rc)
- GOTO(out, retval = ERR_PTR(rc));
+ /* Volatile file name may look like:
+ * <parent>/LUSTRE_VOLATILE_HDR:<mdt_index>:<random>:fd=<fd>
+ * where fd is opened descriptor of reference file.
+ */
+ if (unlikely(filename_is_volatile(dentry->d_name.name,
+ dentry->d_name.len, NULL))) {
+ int ctx_size = LLCRYPT_ENC_CTX_SIZE;
+ struct lustre_sb_info *lsi;
+ struct file *ref_file;
+ struct inode *ref_inode;
+ char *p, *q, *fd_str;
+ void *ctx;
+ int fd;
+
+ p = strnstr(dentry->d_name.name, ":fd=",
+ dentry->d_name.len);
+ if (!p || strlen(p + 4) == 0)
+ GOTO(out, retval = ERR_PTR(-EINVAL));
+
+ q = strchrnul(p + 4, ':');
+ fd_str = kstrndup(p + 4, q - p - 4, GFP_NOFS);
+ if (!fd_str)
+ GOTO(out, retval = ERR_PTR(-ENOMEM));
+ rc = kstrtouint(fd_str, 10, &fd);
+ kfree(fd_str);
+ if (rc)
+ GOTO(inherit, rc = -EINVAL);
+
+ ref_file = fget(fd);
+ if (!ref_file)
+ GOTO(inherit, rc = -EINVAL);
+
+ ref_inode = file_inode(ref_file);
+ if (!ref_inode) {
+ fput(ref_file);
+ GOTO(inherit, rc = -EINVAL);
+ }
+
+ lsi = s2lsi(ref_inode->i_sb);
+
+getctx:
+ OBD_ALLOC(ctx, ctx_size);
+ if (!ctx)
+ GOTO(out, retval = ERR_PTR(-ENOMEM));
+
+#ifdef CONFIG_LL_ENCRYPTION
+ rc = lsi->lsi_cop->get_context(ref_inode,
+ ctx, ctx_size);
+#else
+ rc = -ENODATA;
+#endif
+ if (rc == -ERANGE) {
+ OBD_FREE(ctx, ctx_size);
+ ctx_size *= 2;
+ goto getctx;
+ }
+ fput(ref_file);
+ if (rc < 0) {
+ OBD_FREE(ctx, ctx_size);
+ GOTO(inherit, rc);
+ }
+
+ op_data->op_file_encctx_size = rc;
+ if (rc == ctx_size) {
+ op_data->op_file_encctx = ctx;
+ } else {
+ OBD_ALLOC(op_data->op_file_encctx,
+ op_data->op_file_encctx_size);
+ if (!op_data->op_file_encctx) {
+ OBD_FREE(ctx, ctx_size);
+ GOTO(out, retval = ERR_PTR(-ENOMEM));
+ }
+ memcpy(op_data->op_file_encctx, ctx,
+ op_data->op_file_encctx_size);
+ OBD_FREE(ctx, ctx_size);
+ }
+
+ } else {
+inherit:
+ rc = llcrypt_inherit_context(parent, NULL, op_data,
+ false);
+ if (rc)
+ GOTO(out, retval = ERR_PTR(rc));
+ }
if (encctx != NULL)
*encctx = op_data->op_file_encctx;
if (encctxlen != NULL)
}
if (pca && pca->pca_dataset) {
- struct pcc_dataset *dataset = pca->pca_dataset;
-
OBD_ALLOC_PTR(lum);
if (lum == NULL)
GOTO(out, retval = ERR_PTR(-ENOMEM));
lum->lmm_pattern = LOV_PATTERN_F_RELEASED | LOV_PATTERN_RAID0;
op_data->op_data = lum;
op_data->op_data_size = sizeof(*lum);
- op_data->op_archive_id = dataset->pccd_rwid;
-
- rc = obd_fid_alloc(NULL, ll_i2mdexp(parent), &op_data->op_fid2,
- op_data);
- if (rc)
- GOTO(out, retval = ERR_PTR(rc));
-
- rc = pcc_inode_create(parent->i_sb, dataset, &op_data->op_fid2,
- &pca->pca_dentry);
- if (rc)
- GOTO(out, retval = ERR_PTR(rc));
-
+ op_data->op_archive_id = pca->pca_dataset->pccd_rwid;
it->it_flags |= MDS_OPEN_PCC;
}
if (rc < 0)
GOTO(out, retval = ERR_PTR(rc));
+ if (pca && pca->pca_dataset) {
+ rc = pcc_inode_create(parent->i_sb, pca->pca_dataset,
+ &op_data->op_fid2,
+ &pca->pca_dentry);
+ if (rc)
+ GOTO(out, retval = ERR_PTR(rc));
+ }
+
/* dir layout may change */
ll_unlock_md_op_lsm(op_data);
rc = ll_lookup_it_finish(req, it, parent, &dentry,
op_data->op_file_encctx = NULL;
op_data->op_file_encctx_size = 0;
}
+ llcrypt_free_filename(&fname);
ll_finish_md_op_data(op_data);
}
* we only need to request open lock if it was requested
* for every open
*/
- if (ll_i2sbi(dir)->ll_oc_thrsh_count == 1)
+ if (ll_i2sbi(dir)->ll_oc_thrsh_count == 1 &&
+ exp_connect_flags2(ll_i2mdexp(dir)) &
+ OBD_CONNECT2_ATOMIC_OPEN_LOCK)
it->it_flags |= MDS_OPEN_LOCK;
/* Dentry added to dcache tree in ll_lookup_it */
/* Dentry instantiated in ll_create_it. */
rc = ll_create_it(dir, dentry, it, secctx, secctxlen,
encrypt, encctx, encctxlen);
- security_release_secctx(secctx, secctxlen);
+ ll_security_release_secctx(secctx, secctxlen);
llcrypt_free_ctx(encctx, encctxlen);
if (rc) {
/* We dget in ll_splice_alias. */
LASSERT(it_disposition(it, DISP_ENQ_CREATE_REF));
request = it->it_request;
- it_clear_disposition(it, DISP_ENQ_CREATE_REF);
- rc = ll_prep_inode(&inode, request, dir->i_sb, it);
- if (rc)
- GOTO(out, inode = ERR_PTR(rc));
+ it_clear_disposition(it, DISP_ENQ_CREATE_REF);
+ rc = ll_prep_inode(&inode, &request->rq_pill, dir->i_sb, it);
+ if (rc)
+ GOTO(out, inode = ERR_PTR(rc));
/* Pause to allow for a race with concurrent access by fid */
OBD_FAIL_TIMEOUT(OBD_FAIL_LLITE_CREATE_NODE_PAUSE, cfs_fail_val);
if (IS_ERR(inode))
RETURN(PTR_ERR(inode));
- if ((ll_i2sbi(inode)->ll_flags & LL_SBI_FILE_SECCTX) &&
- secctx != NULL) {
+ if (test_bit(LL_SBI_FILE_SECCTX, ll_i2sbi(inode)->ll_flags) &&
+ secctx) {
/* must be done before d_instantiate, because it calls
* security_d_instantiate, which means a getxattr if security
* context is not set yet */
RETURN(rc);
}
- if (!(ll_i2sbi(inode)->ll_flags & LL_SBI_FILE_SECCTX)) {
+ if (!test_bit(LL_SBI_FILE_SECCTX, ll_i2sbi(inode)->ll_flags)) {
rc = ll_inode_init_security(dentry, inode, dir);
if (rc)
RETURN(rc);
}
ll_set_lock_data(ll_i2sbi(dir)->ll_md_exp, inode, it, &bits);
- if (bits & MDS_INODELOCK_LOOKUP)
+ if (bits & MDS_INODELOCK_LOOKUP) {
d_lustre_revalidate(dentry);
+ ll_update_dir_depth(dir, inode);
+ }
RETURN(0);
}
inode->i_ctime.tv_sec = body->mbo_ctime;
}
+/* once default LMV (space balanced) is set on ROOT, it should take effect if
+ * default LMV is not set on parent directory.
+ */
+static void ll_qos_mkdir_prep(struct md_op_data *op_data, struct inode *dir)
+{
+ struct inode *root = dir->i_sb->s_root->d_inode;
+ struct ll_inode_info *rlli = ll_i2info(root);
+ struct ll_inode_info *lli = ll_i2info(dir);
+ struct lmv_stripe_md *lsm;
+
+ op_data->op_dir_depth = lli->lli_depth;
+
+ /* parent directory is striped */
+ if (unlikely(lli->lli_lsm_md))
+ return;
+
+ /* default LMV set on parent directory */
+ if (unlikely(lli->lli_default_lsm_md))
+ return;
+
+ /* parent is ROOT */
+ if (unlikely(dir == root))
+ return;
+
+ /* default LMV not set on ROOT */
+ if (!rlli->lli_default_lsm_md)
+ return;
+
+ down_read(&rlli->lli_lsm_sem);
+ lsm = rlli->lli_default_lsm_md;
+ if (!lsm)
+ goto unlock;
+
+ /* not space balanced */
+ if (lsm->lsm_md_master_mdt_index != LMV_OFFSET_DEFAULT)
+ goto unlock;
+
+ if (lsm->lsm_md_max_inherit != LMV_INHERIT_NONE &&
+ (lsm->lsm_md_max_inherit == LMV_INHERIT_UNLIMITED ||
+ lsm->lsm_md_max_inherit >= lli->lli_depth)) {
+ op_data->op_flags |= MF_QOS_MKDIR;
+ if (lsm->lsm_md_max_inherit_rr != LMV_INHERIT_RR_NONE &&
+ (lsm->lsm_md_max_inherit_rr == LMV_INHERIT_RR_UNLIMITED ||
+ lsm->lsm_md_max_inherit_rr >= lli->lli_depth))
+ op_data->op_flags |= MF_RR_MKDIR;
+ CDEBUG(D_INODE, DFID" requests qos mkdir %#x\n",
+ PFID(&lli->lli_fid), op_data->op_flags);
+ }
+unlock:
+ up_read(&rlli->lli_lsm_sem);
+}
+
static int ll_new_node(struct inode *dir, struct dentry *dchild,
- const char *tgt, umode_t mode, int rdev, __u32 opc)
+ const char *tgt, umode_t mode, __u64 rdev, __u32 opc)
{
struct qstr *name = &dchild->d_name;
struct ptlrpc_request *request = NULL;
struct md_op_data *op_data = NULL;
struct inode *inode = NULL;
struct ll_sb_info *sbi = ll_i2sbi(dir);
- int tgt_len = 0;
+ struct llcrypt_str *disk_link = NULL;
bool encrypt = false;
int err;
ENTRY;
- if (unlikely(tgt != NULL))
- tgt_len = strlen(tgt) + 1;
+ if (unlikely(tgt != NULL)) {
+ disk_link = (struct llcrypt_str *)rdev;
+ rdev = 0;
+ if (!disk_link)
+ RETURN(-EINVAL);
+ }
again:
op_data = ll_prep_md_op_data(NULL, dir, NULL, name->name,
if (IS_ERR(op_data))
GOTO(err_exit, err = PTR_ERR(op_data));
- if (sbi->ll_flags & LL_SBI_FILE_SECCTX) {
+ if (S_ISDIR(mode))
+ ll_qos_mkdir_prep(op_data, dir);
+
+ if (test_bit(LL_SBI_FILE_SECCTX, sbi->ll_flags)) {
err = ll_dentry_init_security(dchild, mode, &dchild->d_name,
&op_data->op_file_secctx_name,
&op_data->op_file_secctx,
err = llcrypt_inherit_context(dir, NULL, op_data, false);
if (err)
GOTO(err_exit, err);
+
+ if (S_ISLNK(mode)) {
+ /* llcrypt needs inode to encrypt target name, so create
+ * a fake inode and associate encryption context got
+ * from llcrypt_inherit_context.
+ */
+ struct inode *fakeinode =
+ dchild->d_sb->s_op->alloc_inode(dchild->d_sb);
+
+ if (!fakeinode)
+ GOTO(err_exit, err = -ENOMEM);
+ fakeinode->i_sb = dchild->d_sb;
+ fakeinode->i_mode |= S_IFLNK;
+#ifdef IOP_XATTR
+ fakeinode->i_opflags |= IOP_XATTR;
+#endif
+ ll_lli_init(ll_i2info(fakeinode));
+ err = ll_set_encflags(fakeinode,
+ op_data->op_file_encctx,
+ op_data->op_file_encctx_size,
+ true);
+ if (!err)
+ err = __llcrypt_encrypt_symlink(fakeinode, tgt,
+ strlen(tgt),
+ disk_link);
+
+ ll_xattr_cache_destroy(fakeinode);
+ llcrypt_put_encryption_info(fakeinode);
+ dchild->d_sb->s_op->destroy_inode(fakeinode);
+ if (err)
+ GOTO(err_exit, err);
+ }
}
- err = md_create(sbi->ll_md_exp, op_data, tgt, tgt_len, mode,
+ err = md_create(sbi->ll_md_exp, op_data, tgt ? disk_link->name : NULL,
+ tgt ? disk_link->len : 0, mode,
from_kuid(&init_user_ns, current_fsuid()),
from_kgid(&init_user_ns, current_fsgid()),
- cfs_curproc_cap_pack(), rdev, &request);
+ current_cap(), rdev, &request);
#if LUSTRE_VERSION_CODE < OBD_OCD_VERSION(2, 14, 58, 0)
/*
* server < 2.12.58 doesn't pack default LMV in intent_getattr reply,
CFS_FAIL_TIMEOUT(OBD_FAIL_LLITE_NEWNODE_PAUSE, cfs_fail_val);
- err = ll_prep_inode(&inode, request, dchild->d_sb, NULL);
+ err = ll_prep_inode(&inode, &request->rq_pill, dchild->d_sb, NULL);
if (err)
GOTO(err_exit, err);
- if (sbi->ll_flags & LL_SBI_FILE_SECCTX) {
+ if (test_bit(LL_SBI_FILE_SECCTX, sbi->ll_flags)) {
/* must be done before d_instantiate, because it calls
* security_d_instantiate, which means a getxattr if security
* context is not set yet */
op_data->op_file_encctx_size, true);
if (err)
GOTO(err_exit, err);
+
+ if (S_ISLNK(mode)) {
+ struct ll_inode_info *lli = ll_i2info(inode);
+
+ /* Cache the plaintext symlink target
+ * for later use by get_link()
+ */
+ OBD_ALLOC(lli->lli_symlink_name, strlen(tgt) + 1);
+ /* do not return an error if we cannot
+ * cache the symlink locally
+ */
+ if (lli->lli_symlink_name)
+ memcpy(lli->lli_symlink_name,
+ tgt, strlen(tgt) + 1);
+ }
}
- if (!(sbi->ll_flags & LL_SBI_FILE_SECCTX)) {
+ if (!test_bit(LL_SBI_FILE_SECCTX, sbi->ll_flags)) {
err = ll_inode_init_security(dchild, inode, dir);
if (err)
GOTO(err_exit, err);
const char *oldpath)
{
ktime_t kstart = ktime_get();
+ int len = strlen(oldpath);
+ struct llcrypt_str disk_link;
int err;
ENTRY;
CDEBUG(D_VFSTRACE, "VFS Op:name=%pd, dir="DFID"(%p), target=%.*s\n",
dchild, PFID(ll_inode2fid(dir)), dir, 3000, oldpath);
- err = ll_new_node(dir, dchild, oldpath, S_IFLNK | S_IRWXUGO, 0,
- LUSTRE_OPC_SYMLINK);
+ err = llcrypt_prepare_symlink(dir, oldpath, len, dir->i_sb->s_blocksize,
+ &disk_link);
+ if (err)
+ RETURN(err);
+
+ err = ll_new_node(dir, dchild, oldpath, S_IFLNK | S_IRWXUGO,
+ (__u64)&disk_link, LUSTRE_OPC_SYMLINK);
+
+ if (disk_link.name != (unsigned char *)oldpath)
+ kfree(disk_link.name);
if (!err)
ll_stats_ops_tally(ll_i2sbi(dir), LPROC_LL_SYMLINK,
if (dchild->d_inode != NULL)
op_data->op_fid3 = *ll_inode2fid(dchild->d_inode);
- op_data->op_fid2 = op_data->op_fid3;
+ if (fid_is_zero(&op_data->op_fid2))
+ op_data->op_fid2 = op_data->op_fid3;
rc = md_unlink(ll_i2sbi(dir)->ll_md_exp, op_data, &request);
ll_finish_md_op_data(op_data);
if (!rc) {
ll_i2info(dchild->d_inode)->lli_clob &&
dirty_cnt(dchild->d_inode))
op_data->op_cli_flags |= CLI_DIRTY_DATA;
- op_data->op_fid2 = op_data->op_fid3;
+ if (fid_is_zero(&op_data->op_fid2))
+ op_data->op_fid2 = op_data->op_fid3;
rc = md_unlink(ll_i2sbi(dir)->ll_md_exp, op_data, &request);
ll_finish_md_op_data(op_data);
if (rc)
#endif
)
{
- struct qstr *src_name = &src_dchild->d_name;
- struct qstr *tgt_name = &tgt_dchild->d_name;
struct ptlrpc_request *request = NULL;
struct ll_sb_info *sbi = ll_i2sbi(src);
struct md_op_data *op_data;
ktime_t kstart = ktime_get();
umode_t mode = 0;
+ struct llcrypt_name foldname, fnewname;
int err;
ENTRY;
#endif
if (err)
RETURN(err);
+ /* we prevent an encrypted file from being renamed
+ * into an unencrypted dir
+ */
+ if (IS_ENCRYPTED(src) && !IS_ENCRYPTED(tgt))
+ RETURN(-EXDEV);
if (src_dchild->d_inode)
mode = src_dchild->d_inode->i_mode;
if (tgt_dchild->d_inode)
op_data->op_fid4 = *ll_inode2fid(tgt_dchild->d_inode);
+ err = ll_setup_filename(src, &src_dchild->d_name, 1, &foldname, NULL);
+ if (err)
+ RETURN(err);
+ err = ll_setup_filename(tgt, &tgt_dchild->d_name, 1, &fnewname, NULL);
+ if (err) {
+ llcrypt_free_filename(&foldname);
+ RETURN(err);
+ }
err = md_rename(sbi->ll_md_exp, op_data,
- src_name->name, src_name->len,
- tgt_name->name, tgt_name->len, &request);
+ foldname.disk_name.name, foldname.disk_name.len,
+ fnewname.disk_name.name, fnewname.disk_name.len,
+ &request);
+ llcrypt_free_filename(&foldname);
+ llcrypt_free_filename(&fnewname);
ll_finish_md_op_data(op_data);
if (!err) {
ll_update_times(request, src);
git://git.whamcloud.com / fs / lustre-release.git / blobdiff