return 0;
}
-#define CB_SUPPORTED_ATTR0 (FATTR4_WORD0_CHANGE|FATTR4_WORD0_SIZE)
-#define CB_SUPPORTED_ATTR1 (FATTR4_WORD1_TIME_METADATA|FATTR4_WORD1_TIME_MODIFY)
-static __be32 encode_attr_bitmap(struct xdr_stream *xdr, const uint32_t *bitmap, __be32 **savep)
+static __be32 encode_attr_bitmap(struct xdr_stream *xdr, const uint32_t *bitmap, size_t sz)
{
- __be32 bm[2];
- __be32 *p;
-
- bm[0] = htonl(bitmap[0] & CB_SUPPORTED_ATTR0);
- bm[1] = htonl(bitmap[1] & CB_SUPPORTED_ATTR1);
- if (bm[1] != 0) {
- p = xdr_reserve_space(xdr, 16);
- if (unlikely(p == NULL))
- return htonl(NFS4ERR_RESOURCE);
- *p++ = htonl(2);
- *p++ = bm[0];
- *p++ = bm[1];
- } else if (bm[0] != 0) {
- p = xdr_reserve_space(xdr, 12);
- if (unlikely(p == NULL))
- return htonl(NFS4ERR_RESOURCE);
- *p++ = htonl(1);
- *p++ = bm[0];
- } else {
- p = xdr_reserve_space(xdr, 8);
- if (unlikely(p == NULL))
- return htonl(NFS4ERR_RESOURCE);
- *p++ = htonl(0);
- }
- *savep = p;
+ if (xdr_stream_encode_uint32_array(xdr, bitmap, sz) < 0)
+ return cpu_to_be32(NFS4ERR_RESOURCE);
return 0;
}
if (unlikely(status != 0))
goto out;
- status = encode_attr_bitmap(xdr, res->bitmap, &savep);
+ status = encode_attr_bitmap(xdr, res->bitmap, ARRAY_SIZE(res->bitmap));
if (unlikely(status != 0))
goto out;
+ status = cpu_to_be32(NFS4ERR_RESOURCE);
+ savep = xdr_reserve_space(xdr, sizeof(*savep));
+ if (unlikely(!savep))
+ goto out;
status = encode_attr_change(xdr, res->bitmap, res->change_attr);
if (unlikely(status != 0))
goto out;
#include <linux/nfs_xdr.h>
#include "nfs4_fs.h"
+#include "nfs4session.h"
#include "delegation.h"
#include "internal.h"
#include "nfs4trace.h"
* nfs_inode_reclaim_delegation - process a delegation reclaim request
* @inode: inode to process
* @cred: credential to use for request
- * @res: new delegation state from server
+ * @type: delegation type
+ * @stateid: delegation stateid
+ * @pagemod_limit: write delegation "space_limit"
*
*/
void nfs_inode_reclaim_delegation(struct inode *inode, struct rpc_cred *cred,
- struct nfs_openres *res)
+ fmode_t type,
+ const nfs4_stateid *stateid,
+ unsigned long pagemod_limit)
{
struct nfs_delegation *delegation;
struct rpc_cred *oldcred = NULL;
if (delegation != NULL) {
spin_lock(&delegation->lock);
if (delegation->inode != NULL) {
- nfs4_stateid_copy(&delegation->stateid, &res->delegation);
- delegation->type = res->delegation_type;
- delegation->pagemod_limit = res->pagemod_limit;
+ nfs4_stateid_copy(&delegation->stateid, stateid);
+ delegation->type = type;
+ delegation->pagemod_limit = pagemod_limit;
oldcred = delegation->cred;
delegation->cred = get_rpccred(cred);
clear_bit(NFS_DELEGATION_NEED_RECLAIM,
spin_unlock(&delegation->lock);
rcu_read_unlock();
put_rpccred(oldcred);
- trace_nfs4_reclaim_delegation(inode, res->delegation_type);
+ trace_nfs4_reclaim_delegation(inode, type);
return;
}
/* We appear to have raced with a delegation return. */
spin_unlock(&delegation->lock);
}
rcu_read_unlock();
- nfs_inode_set_delegation(inode, cred, res);
+ nfs_inode_set_delegation(inode, cred, type, stateid, pagemod_limit);
}
static int nfs_do_return_delegation(struct inode *inode, struct nfs_delegation *delegation, int issync)
* nfs_inode_set_delegation - set up a delegation on an inode
* @inode: inode to which delegation applies
* @cred: cred to use for subsequent delegation processing
- * @res: new delegation state from server
+ * @type: delegation type
+ * @stateid: delegation stateid
+ * @pagemod_limit: write delegation "space_limit"
*
* Returns zero on success, or a negative errno value.
*/
-int nfs_inode_set_delegation(struct inode *inode, struct rpc_cred *cred, struct nfs_openres *res)
+int nfs_inode_set_delegation(struct inode *inode, struct rpc_cred *cred,
+ fmode_t type,
+ const nfs4_stateid *stateid,
+ unsigned long pagemod_limit)
{
struct nfs_server *server = NFS_SERVER(inode);
struct nfs_client *clp = server->nfs_client;
delegation = kmalloc(sizeof(*delegation), GFP_NOFS);
if (delegation == NULL)
return -ENOMEM;
- nfs4_stateid_copy(&delegation->stateid, &res->delegation);
- delegation->type = res->delegation_type;
- delegation->pagemod_limit = res->pagemod_limit;
+ nfs4_stateid_copy(&delegation->stateid, stateid);
+ delegation->type = type;
+ delegation->pagemod_limit = pagemod_limit;
delegation->change_attr = inode_peek_iversion_raw(inode);
delegation->cred = get_rpccred(cred);
delegation->inode = inode;
rcu_assign_pointer(nfsi->delegation, delegation);
delegation = NULL;
- trace_nfs4_set_delegation(inode, res->delegation_type);
+ trace_nfs4_set_delegation(inode, type);
out:
spin_unlock(&clp->cl_lock);
return err;
}
+/**
+ * nfs4_inode_make_writeable
+ * @inode: pointer to inode
+ *
+ * Make the inode writeable by returning the delegation if necessary
+ *
+ * Returns zero on success, or a negative errno value.
+ */
+int nfs4_inode_make_writeable(struct inode *inode)
+{
+ if (!nfs4_has_session(NFS_SERVER(inode)->nfs_client) ||
+ !nfs4_check_delegation(inode, FMODE_WRITE))
+ return nfs4_inode_return_delegation(inode);
+ return 0;
+}
+
static void nfs_mark_return_if_closed_delegation(struct nfs_server *server,
struct nfs_delegation *delegation)
{
NFS_DELEGATION_TEST_EXPIRED,
};
-int nfs_inode_set_delegation(struct inode *inode, struct rpc_cred *cred, struct nfs_openres *res);
-void nfs_inode_reclaim_delegation(struct inode *inode, struct rpc_cred *cred, struct nfs_openres *res);
+int nfs_inode_set_delegation(struct inode *inode, struct rpc_cred *cred,
+ fmode_t type, const nfs4_stateid *stateid, unsigned long pagemod_limit);
+void nfs_inode_reclaim_delegation(struct inode *inode, struct rpc_cred *cred,
+ fmode_t type, const nfs4_stateid *stateid, unsigned long pagemod_limit);
int nfs4_inode_return_delegation(struct inode *inode);
int nfs_async_inode_return_delegation(struct inode *inode, const nfs4_stateid *stateid);
void nfs_inode_return_delegation_noreclaim(struct inode *inode);
bool nfs4_delegation_flush_on_close(const struct inode *inode);
void nfs_inode_find_delegation_state_and_recover(struct inode *inode,
const nfs4_stateid *stateid);
+int nfs4_inode_make_writeable(struct inode *inode);
#endif
/* drop the inode if we're reasonably sure this is the last link */
if (inode->i_nlink == 1)
clear_nlink(inode);
- NFS_I(inode)->cache_validity |= NFS_INO_INVALID_ATTR;
+ NFS_I(inode)->cache_validity |= NFS_INO_INVALID_CHANGE
+ | NFS_INO_INVALID_CTIME
+ | NFS_INO_INVALID_OTHER;
spin_unlock(&inode->i_lock);
}
trace_nfs_remove_enter(dir, dentry);
if (inode != NULL) {
- NFS_PROTO(inode)->return_delegation(inode);
- error = NFS_PROTO(dir)->remove(dir, &dentry->d_name);
+ error = NFS_PROTO(dir)->remove(dir, dentry);
if (error == 0)
nfs_drop_nlink(inode);
} else
- error = NFS_PROTO(dir)->remove(dir, &dentry->d_name);
+ error = NFS_PROTO(dir)->remove(dir, dentry);
if (error == -ENOENT)
nfs_dentry_handle_enoent(dentry);
trace_nfs_remove_exit(dir, dentry, error);
old_dentry, dentry);
trace_nfs_link_enter(inode, dir, dentry);
- NFS_PROTO(inode)->return_delegation(inode);
-
d_drop(dentry);
error = NFS_PROTO(dir)->link(inode, dir, &dentry->d_name);
if (error == 0) {
}
}
- NFS_PROTO(old_inode)->return_delegation(old_inode);
- if (new_inode != NULL)
- NFS_PROTO(new_inode)->return_delegation(new_inode);
-
task = nfs_async_rename(old_dir, new_dir, old_dentry, new_dentry, NULL);
if (IS_ERR(task)) {
error = PTR_ERR(task);
static void nfs_set_cache_invalid(struct inode *inode, unsigned long flags)
{
struct nfs_inode *nfsi = NFS_I(inode);
+ bool have_delegation = nfs_have_delegated_attributes(inode);
+ if (have_delegation)
+ flags &= ~(NFS_INO_INVALID_CHANGE|NFS_INO_REVAL_PAGECACHE);
if (inode->i_mapping->nrpages == 0)
flags &= ~NFS_INO_INVALID_DATA;
nfsi->cache_validity |= flags;
inode->i_mode = fattr->mode;
if ((fattr->valid & NFS_ATTR_FATTR_MODE) == 0
&& nfs_server_capable(inode, NFS_CAP_MODE))
- nfs_set_cache_invalid(inode, NFS_INO_INVALID_ATTR);
+ nfs_set_cache_invalid(inode, NFS_INO_INVALID_OTHER);
/* Why so? Because we want revalidate for devices/FIFOs, and
* that's precisely what we have in nfs_file_inode_operations.
*/
if (fattr->valid & NFS_ATTR_FATTR_ATIME)
inode->i_atime = fattr->atime;
else if (nfs_server_capable(inode, NFS_CAP_ATIME))
- nfs_set_cache_invalid(inode, NFS_INO_INVALID_ATTR);
+ nfs_set_cache_invalid(inode, NFS_INO_INVALID_ATIME);
if (fattr->valid & NFS_ATTR_FATTR_MTIME)
inode->i_mtime = fattr->mtime;
else if (nfs_server_capable(inode, NFS_CAP_MTIME))
- nfs_set_cache_invalid(inode, NFS_INO_INVALID_ATTR);
+ nfs_set_cache_invalid(inode, NFS_INO_INVALID_MTIME);
if (fattr->valid & NFS_ATTR_FATTR_CTIME)
inode->i_ctime = fattr->ctime;
else if (nfs_server_capable(inode, NFS_CAP_CTIME))
- nfs_set_cache_invalid(inode, NFS_INO_INVALID_ATTR);
+ nfs_set_cache_invalid(inode, NFS_INO_INVALID_CTIME);
if (fattr->valid & NFS_ATTR_FATTR_CHANGE)
inode_set_iversion_raw(inode, fattr->change_attr);
else
- nfs_set_cache_invalid(inode, NFS_INO_INVALID_ATTR
- | NFS_INO_REVAL_PAGECACHE);
+ nfs_set_cache_invalid(inode, NFS_INO_INVALID_CHANGE);
if (fattr->valid & NFS_ATTR_FATTR_SIZE)
inode->i_size = nfs_size_to_loff_t(fattr->size);
else
- nfs_set_cache_invalid(inode, NFS_INO_INVALID_ATTR
- | NFS_INO_REVAL_PAGECACHE);
+ nfs_set_cache_invalid(inode, NFS_INO_INVALID_SIZE);
if (fattr->valid & NFS_ATTR_FATTR_NLINK)
set_nlink(inode, fattr->nlink);
else if (nfs_server_capable(inode, NFS_CAP_NLINK))
- nfs_set_cache_invalid(inode, NFS_INO_INVALID_ATTR);
+ nfs_set_cache_invalid(inode, NFS_INO_INVALID_OTHER);
if (fattr->valid & NFS_ATTR_FATTR_OWNER)
inode->i_uid = fattr->uid;
else if (nfs_server_capable(inode, NFS_CAP_OWNER))
- nfs_set_cache_invalid(inode, NFS_INO_INVALID_ATTR);
+ nfs_set_cache_invalid(inode, NFS_INO_INVALID_OTHER);
if (fattr->valid & NFS_ATTR_FATTR_GROUP)
inode->i_gid = fattr->gid;
else if (nfs_server_capable(inode, NFS_CAP_OWNER_GROUP))
- nfs_set_cache_invalid(inode, NFS_INO_INVALID_ATTR);
+ nfs_set_cache_invalid(inode, NFS_INO_INVALID_OTHER);
if (fattr->valid & NFS_ATTR_FATTR_BLOCKS_USED)
inode->i_blocks = fattr->du.nfs2.blocks;
if (fattr->valid & NFS_ATTR_FATTR_SPACE_USED) {
goto out;
}
- /*
- * Return any delegations if we're going to change ACLs
- */
- if ((attr->ia_valid & (ATTR_MODE|ATTR_UID|ATTR_GID)) != 0)
- NFS_PROTO(inode)->return_delegation(inode);
error = NFS_PROTO(inode)->setattr(dentry, fattr, attr);
if (error == 0)
error = nfs_refresh_inode(inode, fattr);
/* Optimisation */
if (offset == 0)
NFS_I(inode)->cache_validity &= ~NFS_INO_INVALID_DATA;
+ NFS_I(inode)->cache_validity &= ~NFS_INO_INVALID_SIZE;
spin_unlock(&inode->i_lock);
truncate_pagecache(inode, offset);
* nfs_setattr_update_inode - Update inode metadata after a setattr call.
* @inode: pointer to struct inode
* @attr: pointer to struct iattr
+ * @fattr: pointer to struct nfs_fattr
*
* Note: we do this in the *proc.c in order to ensure that
* it works for things like exclusive creates too.
spin_lock(&inode->i_lock);
NFS_I(inode)->attr_gencount = fattr->gencount;
+ nfs_set_cache_invalid(inode, NFS_INO_INVALID_CHANGE
+ | NFS_INO_INVALID_CTIME);
if ((attr->ia_valid & (ATTR_MODE|ATTR_UID|ATTR_GID)) != 0) {
if ((attr->ia_valid & ATTR_MODE) != 0) {
int mode = attr->ia_mode & S_IALLUGO;
| NFS_INO_INVALID_ACL);
}
if ((attr->ia_valid & ATTR_SIZE) != 0) {
+ nfs_set_cache_invalid(inode, NFS_INO_INVALID_MTIME);
nfs_inc_stats(inode, NFSIOS_SETATTRTRUNC);
nfs_vmtruncate(inode, attr->ia_size);
}
if (fattr->valid)
nfs_update_inode(inode, fattr);
- else
- NFS_I(inode)->cache_validity |= NFS_INO_INVALID_ATTR;
spin_unlock(&inode->i_lock);
}
EXPORT_SYMBOL_GPL(nfs_setattr_update_inode);
return nfs_file_has_writers(nfsi) && nfs_file_io_is_buffered(nfsi);
}
-static unsigned long nfs_wcc_update_inode(struct inode *inode, struct nfs_fattr *fattr)
+static void nfs_wcc_update_inode(struct inode *inode, struct nfs_fattr *fattr)
{
- unsigned long ret = 0;
-
if ((fattr->valid & NFS_ATTR_FATTR_PRECHANGE)
&& (fattr->valid & NFS_ATTR_FATTR_CHANGE)
&& inode_eq_iversion_raw(inode, fattr->pre_change_attr)) {
inode_set_iversion_raw(inode, fattr->change_attr);
if (S_ISDIR(inode->i_mode))
nfs_set_cache_invalid(inode, NFS_INO_INVALID_DATA);
- ret |= NFS_INO_INVALID_ATTR;
}
/* If we have atomic WCC data, we may update some attributes */
if ((fattr->valid & NFS_ATTR_FATTR_PRECTIME)
&& (fattr->valid & NFS_ATTR_FATTR_CTIME)
&& timespec_equal(&inode->i_ctime, &fattr->pre_ctime)) {
memcpy(&inode->i_ctime, &fattr->ctime, sizeof(inode->i_ctime));
- ret |= NFS_INO_INVALID_ATTR;
}
if ((fattr->valid & NFS_ATTR_FATTR_PREMTIME)
memcpy(&inode->i_mtime, &fattr->mtime, sizeof(inode->i_mtime));
if (S_ISDIR(inode->i_mode))
nfs_set_cache_invalid(inode, NFS_INO_INVALID_DATA);
- ret |= NFS_INO_INVALID_ATTR;
}
if ((fattr->valid & NFS_ATTR_FATTR_PRESIZE)
&& (fattr->valid & NFS_ATTR_FATTR_SIZE)
&& i_size_read(inode) == nfs_size_to_loff_t(fattr->pre_size)
&& !nfs_have_writebacks(inode)) {
i_size_write(inode, nfs_size_to_loff_t(fattr->size));
- ret |= NFS_INO_INVALID_ATTR;
}
-
- return ret;
}
/**
if (!nfs_file_has_buffered_writers(nfsi)) {
/* Verify a few of the more important attributes */
if ((fattr->valid & NFS_ATTR_FATTR_CHANGE) != 0 && !inode_eq_iversion_raw(inode, fattr->change_attr))
- invalid |= NFS_INO_INVALID_ATTR | NFS_INO_REVAL_PAGECACHE;
+ invalid |= NFS_INO_INVALID_CHANGE
+ | NFS_INO_REVAL_PAGECACHE;
if ((fattr->valid & NFS_ATTR_FATTR_MTIME) && !timespec_equal(&inode->i_mtime, &fattr->mtime))
- invalid |= NFS_INO_INVALID_ATTR;
+ invalid |= NFS_INO_INVALID_MTIME;
if ((fattr->valid & NFS_ATTR_FATTR_CTIME) && !timespec_equal(&inode->i_ctime, &fattr->ctime))
- invalid |= NFS_INO_INVALID_ATTR;
+ invalid |= NFS_INO_INVALID_CTIME;
if (fattr->valid & NFS_ATTR_FATTR_SIZE) {
cur_size = i_size_read(inode);
new_isize = nfs_size_to_loff_t(fattr->size);
if (cur_size != new_isize)
- invalid |= NFS_INO_INVALID_ATTR|NFS_INO_REVAL_PAGECACHE;
+ invalid |= NFS_INO_INVALID_SIZE
+ | NFS_INO_REVAL_PAGECACHE;
}
}
/* Have any file permissions changed? */
if ((fattr->valid & NFS_ATTR_FATTR_MODE) && (inode->i_mode & S_IALLUGO) != (fattr->mode & S_IALLUGO))
- invalid |= NFS_INO_INVALID_ATTR | NFS_INO_INVALID_ACCESS | NFS_INO_INVALID_ACL;
+ invalid |= NFS_INO_INVALID_ACCESS
+ | NFS_INO_INVALID_ACL
+ | NFS_INO_INVALID_OTHER;
if ((fattr->valid & NFS_ATTR_FATTR_OWNER) && !uid_eq(inode->i_uid, fattr->uid))
- invalid |= NFS_INO_INVALID_ATTR | NFS_INO_INVALID_ACCESS | NFS_INO_INVALID_ACL;
+ invalid |= NFS_INO_INVALID_ACCESS
+ | NFS_INO_INVALID_ACL
+ | NFS_INO_INVALID_OTHER;
if ((fattr->valid & NFS_ATTR_FATTR_GROUP) && !gid_eq(inode->i_gid, fattr->gid))
- invalid |= NFS_INO_INVALID_ATTR | NFS_INO_INVALID_ACCESS | NFS_INO_INVALID_ACL;
+ invalid |= NFS_INO_INVALID_ACCESS
+ | NFS_INO_INVALID_ACL
+ | NFS_INO_INVALID_OTHER;
/* Has the link count changed? */
if ((fattr->valid & NFS_ATTR_FATTR_NLINK) && inode->i_nlink != fattr->nlink)
- invalid |= NFS_INO_INVALID_ATTR;
+ invalid |= NFS_INO_INVALID_OTHER;
if ((fattr->valid & NFS_ATTR_FATTR_ATIME) && !timespec_equal(&inode->i_atime, &fattr->atime))
invalid |= NFS_INO_INVALID_ATIME;
}
EXPORT_SYMBOL_GPL(nfs_refresh_inode);
-static int nfs_post_op_update_inode_locked(struct inode *inode, struct nfs_fattr *fattr)
+static int nfs_post_op_update_inode_locked(struct inode *inode,
+ struct nfs_fattr *fattr, unsigned int invalid)
{
- unsigned long invalid = NFS_INO_INVALID_ATTR;
-
if (S_ISDIR(inode->i_mode))
invalid |= NFS_INO_INVALID_DATA;
nfs_set_cache_invalid(inode, invalid);
spin_lock(&inode->i_lock);
nfs_fattr_set_barrier(fattr);
- status = nfs_post_op_update_inode_locked(inode, fattr);
+ status = nfs_post_op_update_inode_locked(inode, fattr,
+ NFS_INO_INVALID_CHANGE
+ | NFS_INO_INVALID_CTIME);
spin_unlock(&inode->i_lock);
return status;
fattr->valid |= NFS_ATTR_FATTR_PRESIZE;
}
out_noforce:
- status = nfs_post_op_update_inode_locked(inode, fattr);
+ status = nfs_post_op_update_inode_locked(inode, fattr,
+ NFS_INO_INVALID_CHANGE
+ | NFS_INO_INVALID_CTIME
+ | NFS_INO_INVALID_MTIME);
return status;
}
| NFS_INO_REVAL_PAGECACHE);
/* Do atomic weak cache consistency updates */
- invalid |= nfs_wcc_update_inode(inode, fattr);
+ nfs_wcc_update_inode(inode, fattr);
if (pnfs_layoutcommit_outstanding(inode)) {
nfsi->cache_validity |= save_cache_validity & NFS_INO_INVALID_ATTR;
inode->i_sb->s_id, inode->i_ino);
/* Could it be a race with writeback? */
if (!have_writers) {
- invalid |= NFS_INO_INVALID_ATTR
+ invalid |= NFS_INO_INVALID_CHANGE
| NFS_INO_INVALID_DATA
| NFS_INO_INVALID_ACCESS
| NFS_INO_INVALID_ACL;
+ /* Force revalidate of all attributes */
+ save_cache_validity |= NFS_INO_INVALID_CTIME
+ | NFS_INO_INVALID_MTIME
+ | NFS_INO_INVALID_SIZE
+ | NFS_INO_INVALID_OTHER;
if (S_ISDIR(inode->i_mode))
nfs_force_lookup_revalidate(inode);
}
inode_set_iversion_raw(inode, fattr->change_attr);
}
} else {
- nfsi->cache_validity |= save_cache_validity;
+ nfsi->cache_validity |= save_cache_validity &
+ (NFS_INO_INVALID_CHANGE
+ | NFS_INO_REVAL_PAGECACHE
+ | NFS_INO_REVAL_FORCED);
cache_revalidated = false;
}
memcpy(&inode->i_mtime, &fattr->mtime, sizeof(inode->i_mtime));
} else if (server->caps & NFS_CAP_MTIME) {
nfsi->cache_validity |= save_cache_validity &
- (NFS_INO_INVALID_ATTR
+ (NFS_INO_INVALID_MTIME
| NFS_INO_REVAL_FORCED);
cache_revalidated = false;
}
memcpy(&inode->i_ctime, &fattr->ctime, sizeof(inode->i_ctime));
} else if (server->caps & NFS_CAP_CTIME) {
nfsi->cache_validity |= save_cache_validity &
- (NFS_INO_INVALID_ATTR
+ (NFS_INO_INVALID_CTIME
| NFS_INO_REVAL_FORCED);
cache_revalidated = false;
}
if (!nfs_have_writebacks(inode) || new_isize > cur_isize) {
i_size_write(inode, new_isize);
if (!have_writers)
- invalid |= NFS_INO_INVALID_ATTR|NFS_INO_INVALID_DATA;
+ invalid |= NFS_INO_INVALID_DATA;
}
dprintk("NFS: isize change on server for file %s/%ld "
"(%Ld to %Ld)\n",
}
} else {
nfsi->cache_validity |= save_cache_validity &
- (NFS_INO_INVALID_ATTR
+ (NFS_INO_INVALID_SIZE
| NFS_INO_REVAL_PAGECACHE
| NFS_INO_REVAL_FORCED);
cache_revalidated = false;
umode_t newmode = inode->i_mode & S_IFMT;
newmode |= fattr->mode & S_IALLUGO;
inode->i_mode = newmode;
- invalid |= NFS_INO_INVALID_ATTR|NFS_INO_INVALID_ACCESS|NFS_INO_INVALID_ACL;
+ invalid |= NFS_INO_INVALID_ACCESS
+ | NFS_INO_INVALID_ACL
+ | NFS_INO_INVALID_OTHER;
}
} else if (server->caps & NFS_CAP_MODE) {
nfsi->cache_validity |= save_cache_validity &
- (NFS_INO_INVALID_ATTR
- | NFS_INO_INVALID_ACCESS
+ (NFS_INO_INVALID_ACCESS
| NFS_INO_INVALID_ACL
+ | NFS_INO_INVALID_OTHER
| NFS_INO_REVAL_FORCED);
cache_revalidated = false;
}
if (fattr->valid & NFS_ATTR_FATTR_OWNER) {
if (!uid_eq(inode->i_uid, fattr->uid)) {
- invalid |= NFS_INO_INVALID_ATTR|NFS_INO_INVALID_ACCESS|NFS_INO_INVALID_ACL;
+ invalid |= NFS_INO_INVALID_ACCESS
+ | NFS_INO_INVALID_ACL
+ | NFS_INO_INVALID_OTHER;
inode->i_uid = fattr->uid;
}
} else if (server->caps & NFS_CAP_OWNER) {
nfsi->cache_validity |= save_cache_validity &
- (NFS_INO_INVALID_ATTR
- | NFS_INO_INVALID_ACCESS
+ (NFS_INO_INVALID_ACCESS
| NFS_INO_INVALID_ACL
+ | NFS_INO_INVALID_OTHER
| NFS_INO_REVAL_FORCED);
cache_revalidated = false;
}
if (fattr->valid & NFS_ATTR_FATTR_GROUP) {
if (!gid_eq(inode->i_gid, fattr->gid)) {
- invalid |= NFS_INO_INVALID_ATTR|NFS_INO_INVALID_ACCESS|NFS_INO_INVALID_ACL;
+ invalid |= NFS_INO_INVALID_ACCESS
+ | NFS_INO_INVALID_ACL
+ | NFS_INO_INVALID_OTHER;
inode->i_gid = fattr->gid;
}
} else if (server->caps & NFS_CAP_OWNER_GROUP) {
nfsi->cache_validity |= save_cache_validity &
- (NFS_INO_INVALID_ATTR
- | NFS_INO_INVALID_ACCESS
+ (NFS_INO_INVALID_ACCESS
| NFS_INO_INVALID_ACL
+ | NFS_INO_INVALID_OTHER
| NFS_INO_REVAL_FORCED);
cache_revalidated = false;
}
if (fattr->valid & NFS_ATTR_FATTR_NLINK) {
if (inode->i_nlink != fattr->nlink) {
- invalid |= NFS_INO_INVALID_ATTR;
+ invalid |= NFS_INO_INVALID_OTHER;
if (S_ISDIR(inode->i_mode))
invalid |= NFS_INO_INVALID_DATA;
set_nlink(inode, fattr->nlink);
}
} else if (server->caps & NFS_CAP_NLINK) {
nfsi->cache_validity |= save_cache_validity &
- (NFS_INO_INVALID_ATTR
+ (NFS_INO_INVALID_OTHER
| NFS_INO_REVAL_FORCED);
cache_revalidated = false;
}
/* Update attrtimeo value if we're out of the unstable period */
if (invalid & NFS_INO_INVALID_ATTR) {
+ invalid &= ~NFS_INO_INVALID_ATTR;
nfs_inc_stats(inode, NFSIOS_ATTRINVALIDATE);
nfsi->attrtimeo = NFS_MINATTRTIMEO(inode);
nfsi->attrtimeo_timestamp = now;
nfsi->attr_gencount = fattr->gencount;
}
- /* Don't declare attrcache up to date if there were no attrs! */
- if (cache_revalidated)
- invalid &= ~NFS_INO_INVALID_ATTR;
-
/* Don't invalidate the data if we were to blame */
if (!(S_ISREG(inode->i_mode) || S_ISDIR(inode->i_mode)
|| S_ISLNK(inode->i_mode)))
msg.rpc_cred = nfs_file_cred(sattr->ia_file);
nfs_fattr_init(fattr);
status = rpc_call_sync(NFS_CLIENT(inode), &msg, 0);
- if (status == 0)
+ if (status == 0) {
+ if (NFS_I(inode)->cache_validity & NFS_INO_INVALID_ACL)
+ nfs_zap_acl_cache(inode);
nfs_setattr_update_inode(inode, sattr, fattr);
+ }
dprintk("NFS reply setattr: %d\n", status);
return status;
}
}
static int
-nfs3_proc_remove(struct inode *dir, const struct qstr *name)
+nfs3_proc_remove(struct inode *dir, struct dentry *dentry)
{
struct nfs_removeargs arg = {
.fh = NFS_FH(dir),
- .name = *name,
+ .name = dentry->d_name,
};
struct nfs_removeres res;
struct rpc_message msg = {
};
int status = -ENOMEM;
- dprintk("NFS call remove %s\n", name->name);
+ dprintk("NFS call remove %pd2\n", dentry);
res.dir_attr = nfs_alloc_fattr();
if (res.dir_attr == NULL)
goto out;
}
static void
-nfs3_proc_unlink_setup(struct rpc_message *msg, struct inode *dir)
+nfs3_proc_unlink_setup(struct rpc_message *msg, struct dentry *dentry)
{
msg->rpc_proc = &nfs3_procedures[NFS3PROC_REMOVE];
}
}
static void
-nfs3_proc_rename_setup(struct rpc_message *msg, struct inode *dir)
+nfs3_proc_rename_setup(struct rpc_message *msg,
+ struct dentry *old_dentry,
+ struct dentry *new_dentry)
{
msg->rpc_proc = &nfs3_procedures[NFS3PROC_RENAME];
}
return 0;
}
-static int nfs3_return_delegation(struct inode *inode)
-{
- nfs_wb_all(inode);
- return 0;
-}
-
static const struct inode_operations nfs3_dir_inode_operations = {
.create = nfs_create,
.lookup = nfs_lookup,
.clear_acl_cache = forget_all_cached_acls,
.close_context = nfs_close_context,
.have_delegation = nfs3_have_delegation,
- .return_delegation = nfs3_return_delegation,
.alloc_client = nfs_alloc_client,
.init_client = nfs_init_client,
.free_client = nfs_free_client,
struct nfs_entry old = *entry;
__be32 *p;
int error;
+ u64 new_cookie;
p = xdr_inline_decode(xdr, 4);
if (unlikely(p == NULL))
if (unlikely(error))
return error;
- entry->prev_cookie = entry->cookie;
- error = decode_cookie3(xdr, &entry->cookie);
+ error = decode_cookie3(xdr, &new_cookie);
if (unlikely(error))
return error;
zero_nfs_fh3(entry->fh);
}
+ entry->prev_cookie = entry->cookie;
+ entry->cookie = new_cookie;
+
return 0;
out_overflow:
struct nfs_inode *nfsi = NFS_I(dir);
spin_lock(&dir->i_lock);
- nfsi->cache_validity |= NFS_INO_INVALID_ATTR|NFS_INO_INVALID_DATA;
+ nfsi->cache_validity |= NFS_INO_INVALID_CTIME
+ | NFS_INO_INVALID_MTIME
+ | NFS_INO_INVALID_DATA;
if (cinfo->atomic && cinfo->before == inode_peek_iversion_raw(dir)) {
nfsi->cache_validity &= ~NFS_INO_REVAL_PAGECACHE;
nfsi->attrtimeo_timestamp = jiffies;
{
struct nfs_delegation *delegation;
+ fmode &= FMODE_READ|FMODE_WRITE;
rcu_read_lock();
delegation = rcu_dereference(NFS_I(inode)->delegation);
if (delegation == NULL || (delegation->type & fmode) == fmode) {
}
if ((delegation_flags & 1UL<<NFS_DELEGATION_NEED_RECLAIM) == 0)
nfs_inode_set_delegation(state->inode,
- data->owner->so_cred,
- &data->o_res);
+ data->owner->so_cred,
+ data->o_res.delegation_type,
+ &data->o_res.delegation,
+ data->o_res.pagemod_limit);
else
nfs_inode_reclaim_delegation(state->inode,
- data->owner->so_cred,
- &data->o_res);
+ data->owner->so_cred,
+ data->o_res.delegation_type,
+ &data->o_res.delegation,
+ data->o_res.pagemod_limit);
}
/*
* fields corresponding to attributes that were used to store the verifier.
* Make sure we clobber those fields in the later setattr call
*/
-static inline void nfs4_exclusive_attrset(struct nfs4_opendata *opendata,
+static unsigned nfs4_exclusive_attrset(struct nfs4_opendata *opendata,
struct iattr *sattr, struct nfs4_label **label)
{
- const u32 *attrset = opendata->o_res.attrset;
+ const __u32 *bitmask = opendata->o_arg.server->exclcreat_bitmask;
+ __u32 attrset[3];
+ unsigned ret;
+ unsigned i;
- if ((attrset[1] & FATTR4_WORD1_TIME_ACCESS) &&
- !(sattr->ia_valid & ATTR_ATIME_SET))
- sattr->ia_valid |= ATTR_ATIME;
+ for (i = 0; i < ARRAY_SIZE(attrset); i++) {
+ attrset[i] = opendata->o_res.attrset[i];
+ if (opendata->o_arg.createmode == NFS4_CREATE_EXCLUSIVE4_1)
+ attrset[i] &= ~bitmask[i];
+ }
+
+ ret = (opendata->o_arg.createmode == NFS4_CREATE_EXCLUSIVE) ?
+ sattr->ia_valid : 0;
- if ((attrset[1] & FATTR4_WORD1_TIME_MODIFY) &&
- !(sattr->ia_valid & ATTR_MTIME_SET))
- sattr->ia_valid |= ATTR_MTIME;
+ if ((attrset[1] & (FATTR4_WORD1_TIME_ACCESS|FATTR4_WORD1_TIME_ACCESS_SET))) {
+ if (sattr->ia_valid & ATTR_ATIME_SET)
+ ret |= ATTR_ATIME_SET;
+ else
+ ret |= ATTR_ATIME;
+ }
- /* Except MODE, it seems harmless of setting twice. */
- if (opendata->o_arg.createmode != NFS4_CREATE_EXCLUSIVE &&
- (attrset[1] & FATTR4_WORD1_MODE ||
- attrset[2] & FATTR4_WORD2_MODE_UMASK))
- sattr->ia_valid &= ~ATTR_MODE;
+ if ((attrset[1] & (FATTR4_WORD1_TIME_MODIFY|FATTR4_WORD1_TIME_MODIFY_SET))) {
+ if (sattr->ia_valid & ATTR_MTIME_SET)
+ ret |= ATTR_MTIME_SET;
+ else
+ ret |= ATTR_MTIME;
+ }
- if (attrset[2] & FATTR4_WORD2_SECURITY_LABEL)
+ if (!(attrset[2] & FATTR4_WORD2_SECURITY_LABEL))
*label = NULL;
+ return ret;
}
static int _nfs4_open_and_get_state(struct nfs4_opendata *opendata,
if ((opendata->o_arg.open_flags & (O_CREAT|O_EXCL)) == (O_CREAT|O_EXCL) &&
(opendata->o_arg.createmode != NFS4_CREATE_GUARDED)) {
- nfs4_exclusive_attrset(opendata, sattr, &label);
+ unsigned attrs = nfs4_exclusive_attrset(opendata, sattr, &label);
/*
* send create attributes which was not set by open
* with an extra setattr.
*/
- if (sattr->ia_valid & NFS4_VALID_ATTRS) {
+ if (attrs || label) {
+ unsigned ia_old = sattr->ia_valid;
+
+ sattr->ia_valid = attrs;
nfs_fattr_init(opendata->o_res.f_attr);
status = nfs4_do_setattr(state->inode, cred,
opendata->o_res.f_attr, sattr,
opendata->o_res.f_attr);
nfs_setsecurity(state->inode, opendata->o_res.f_attr, olabel);
}
+ sattr->ia_valid = ia_old;
}
}
if (opened && opendata->file_created)
if (IS_ERR(label))
return PTR_ERR(label);
+ /* Return any delegations if we're going to change ACLs */
+ if ((sattr->ia_valid & (ATTR_MODE|ATTR_UID|ATTR_GID)) != 0)
+ nfs4_inode_make_writeable(inode);
+
status = nfs4_do_setattr(inode, cred, fattr, sattr, ctx, NULL, label);
if (status == 0) {
nfs_setattr_update_inode(inode, sattr, fattr);
struct nfs_server *server = NFS_SERVER(inode);
struct nfs4_accessargs args = {
.fh = NFS_FH(inode),
- .bitmask = server->cache_consistency_bitmask,
.access = entry->mask,
};
struct nfs4_accessres res = {
};
int status = 0;
- res.fattr = nfs_alloc_fattr();
- if (res.fattr == NULL)
- return -ENOMEM;
+ if (!nfs_have_delegated_attributes(inode)) {
+ res.fattr = nfs_alloc_fattr();
+ if (res.fattr == NULL)
+ return -ENOMEM;
+ args.bitmask = server->cache_consistency_bitmask;
+ }
status = nfs4_call_sync(server->client, server, &msg, &args.seq_args, &res.seq_res, 0);
if (!status) {
nfs_access_set_mask(entry, res.access);
- nfs_refresh_inode(inode, res.fattr);
+ if (res.fattr)
+ nfs_refresh_inode(inode, res.fattr);
}
nfs_free_fattr(res.fattr);
return status;
return status;
}
-static int nfs4_proc_remove(struct inode *dir, const struct qstr *name)
+static int nfs4_proc_remove(struct inode *dir, struct dentry *dentry)
+{
+ struct nfs4_exception exception = { };
+ struct inode *inode = d_inode(dentry);
+ int err;
+
+ if (inode) {
+ if (inode->i_nlink == 1)
+ nfs4_inode_return_delegation(inode);
+ else
+ nfs4_inode_make_writeable(inode);
+ }
+ do {
+ err = _nfs4_proc_remove(dir, &dentry->d_name);
+ trace_nfs4_remove(dir, &dentry->d_name, err);
+ err = nfs4_handle_exception(NFS_SERVER(dir), err,
+ &exception);
+ } while (exception.retry);
+ return err;
+}
+
+static int nfs4_proc_rmdir(struct inode *dir, const struct qstr *name)
{
struct nfs4_exception exception = { };
int err;
+
do {
err = _nfs4_proc_remove(dir, name);
trace_nfs4_remove(dir, name, err);
return err;
}
-static void nfs4_proc_unlink_setup(struct rpc_message *msg, struct inode *dir)
+static void nfs4_proc_unlink_setup(struct rpc_message *msg, struct dentry *dentry)
{
- struct nfs_server *server = NFS_SERVER(dir);
struct nfs_removeargs *args = msg->rpc_argp;
struct nfs_removeres *res = msg->rpc_resp;
+ struct inode *inode = d_inode(dentry);
- res->server = server;
+ res->server = NFS_SB(dentry->d_sb);
msg->rpc_proc = &nfs4_procedures[NFSPROC4_CLNT_REMOVE];
nfs4_init_sequence(&args->seq_args, &res->seq_res, 1);
nfs_fattr_init(res->dir_attr);
+
+ if (inode)
+ nfs4_inode_return_delegation(inode);
}
static void nfs4_proc_unlink_rpc_prepare(struct rpc_task *task, struct nfs_unlinkdata *data)
return 1;
}
-static void nfs4_proc_rename_setup(struct rpc_message *msg, struct inode *dir)
+static void nfs4_proc_rename_setup(struct rpc_message *msg,
+ struct dentry *old_dentry,
+ struct dentry *new_dentry)
{
- struct nfs_server *server = NFS_SERVER(dir);
struct nfs_renameargs *arg = msg->rpc_argp;
struct nfs_renameres *res = msg->rpc_resp;
+ struct inode *old_inode = d_inode(old_dentry);
+ struct inode *new_inode = d_inode(new_dentry);
+ if (old_inode)
+ nfs4_inode_make_writeable(old_inode);
+ if (new_inode)
+ nfs4_inode_return_delegation(new_inode);
msg->rpc_proc = &nfs4_procedures[NFSPROC4_CLNT_RENAME];
- res->server = server;
+ res->server = NFS_SB(old_dentry->d_sb);
nfs4_init_sequence(&arg->seq_args, &res->seq_res, 1);
}
}
arg.bitmask = nfs4_bitmask(server, res.label);
+ nfs4_inode_make_writeable(inode);
+
status = nfs4_call_sync(server->client, server, &msg, &arg.seq_args, &res.seq_res, 1);
if (!status) {
update_changeattr(dir, &res.cinfo, res.fattr->time_start);
i = buf_to_pages_noslab(buf, buflen, arg.acl_pages);
if (i < 0)
return i;
- nfs4_inode_return_delegation(inode);
+ nfs4_inode_make_writeable(inode);
ret = nfs4_call_sync(server->client, server, &msg, &arg.seq_args, &res.seq_res, 1);
/*
* so mark the attribute cache invalid.
*/
spin_lock(&inode->i_lock);
- NFS_I(inode)->cache_validity |= NFS_INO_INVALID_ATTR;
+ NFS_I(inode)->cache_validity |= NFS_INO_INVALID_CHANGE
+ | NFS_INO_INVALID_CTIME;
spin_unlock(&inode->i_lock);
nfs_access_zap_cache(inode);
nfs_zap_acl_cache(inode);
nfs4_wake_lock_waiter(wait_queue_entry_t *wait, unsigned int mode, int flags, void *key)
{
int ret;
- struct cb_notify_lock_args *cbnl = key;
struct nfs4_lock_waiter *waiter = wait->private;
- struct nfs_lowner *lowner = &cbnl->cbnl_owner,
- *wowner = waiter->owner;
- /* Only wake if the callback was for the same owner */
- if (lowner->clientid != wowner->clientid ||
- lowner->id != wowner->id ||
- lowner->s_dev != wowner->s_dev)
- return 0;
+ /* NULL key means to wake up everyone */
+ if (key) {
+ struct cb_notify_lock_args *cbnl = key;
+ struct nfs_lowner *lowner = &cbnl->cbnl_owner,
+ *wowner = waiter->owner;
- /* Make sure it's for the right inode */
- if (nfs_compare_fh(NFS_FH(waiter->inode), &cbnl->cbnl_fh))
- return 0;
+ /* Only wake if the callback was for the same owner. */
+ if (lowner->id != wowner->id || lowner->s_dev != wowner->s_dev)
+ return 0;
- waiter->notified = true;
+ /* Make sure it's for the right inode */
+ if (nfs_compare_fh(NFS_FH(waiter->inode), &cbnl->cbnl_fh))
+ return 0;
+
+ waiter->notified = true;
+ }
/* override "private" so we can use default_wake_function */
wait->private = waiter->task;
add_wait_queue(q, &wait);
while(!signalled()) {
+ waiter.notified = false;
status = nfs4_proc_setlk(state, cmd, request);
if ((status != -EAGAIN) || IS_SETLK(cmd))
break;
{
switch(task->tk_status) {
case 0:
+ wake_up_all(&clp->cl_lock_waitq);
+ /* Fallthrough */
case -NFS4ERR_COMPLETE_ALREADY:
case -NFS4ERR_WRONG_CRED: /* What to do here? */
break;
.link = nfs4_proc_link,
.symlink = nfs4_proc_symlink,
.mkdir = nfs4_proc_mkdir,
- .rmdir = nfs4_proc_remove,
+ .rmdir = nfs4_proc_rmdir,
.readdir = nfs4_proc_readdir,
.mknod = nfs4_proc_mknod,
.statfs = nfs4_proc_statfs,
.close_context = nfs4_close_context,
.open_context = nfs4_atomic_open,
.have_delegation = nfs4_have_delegation,
- .return_delegation = nfs4_inode_return_delegation,
.alloc_client = nfs4_alloc_client,
.init_client = nfs4_init_client,
.free_client = nfs4_free_client,
struct rb_node **p = &server->state_owners.rb_node,
*parent = NULL;
struct nfs4_state_owner *sp;
- int err;
while (*p != NULL) {
parent = *p;
return sp;
}
}
- err = ida_get_new(&server->openowner_id, &new->so_seqid.owner_id);
- if (err)
- return ERR_PTR(err);
rb_link_node(&new->so_server_node, parent, p);
rb_insert_color(&new->so_server_node, &server->state_owners);
return new;
if (!RB_EMPTY_NODE(&sp->so_server_node))
rb_erase(&sp->so_server_node, &server->state_owners);
- ida_remove(&server->openowner_id, sp->so_seqid.owner_id);
}
static void
sp = kzalloc(sizeof(*sp), gfp_flags);
if (!sp)
return NULL;
+ sp->so_seqid.owner_id = ida_simple_get(&server->openowner_id, 0, 0,
+ gfp_flags);
+ if (sp->so_seqid.owner_id < 0) {
+ kfree(sp);
+ return NULL;
+ }
sp->so_server = server;
sp->so_cred = get_rpccred(cred);
spin_lock_init(&sp->so_lock);
{
nfs4_destroy_seqid_counter(&sp->so_seqid);
put_rpccred(sp->so_cred);
+ ida_simple_remove(&sp->so_server->openowner_id, sp->so_seqid.owner_id);
kfree(sp);
}
new = nfs4_alloc_state_owner(server, cred, gfp_flags);
if (new == NULL)
goto out;
- do {
- if (ida_pre_get(&server->openowner_id, gfp_flags) == 0)
- break;
- spin_lock(&clp->cl_lock);
- sp = nfs4_insert_state_owner_locked(new);
- spin_unlock(&clp->cl_lock);
- } while (sp == ERR_PTR(-EAGAIN));
+ spin_lock(&clp->cl_lock);
+ sp = nfs4_insert_state_owner_locked(new);
+ spin_unlock(&clp->cl_lock);
if (sp != new)
nfs4_free_state_owner(new);
out:
((3+NFS4_FHSIZE) >> 2))
#define nfs4_fattr_bitmap_maxsz 4
#define encode_getattr_maxsz (op_encode_hdr_maxsz + nfs4_fattr_bitmap_maxsz)
+#define nfstime4_maxsz (3)
#define nfs4_name_maxsz (1 + ((3 + NFS4_MAXNAMLEN) >> 2))
#define nfs4_path_maxsz (1 + ((3 + NFS4_MAXPATHLEN) >> 2))
#define nfs4_owner_maxsz (1 + XDR_QUADLEN(IDMAP_NAMESZ))
#define decode_mdsthreshold_maxsz (1 + 1 + nfs4_fattr_bitmap_maxsz + 1 + 8)
/* This is based on getfattr, which uses the most attributes: */
#define nfs4_fattr_value_maxsz (1 + (1 + 2 + 2 + 4 + 2 + 1 + 1 + 2 + 2 + \
- 3 + 3 + 3 + nfs4_owner_maxsz + \
+ 3*nfstime4_maxsz + \
+ nfs4_owner_maxsz + \
nfs4_group_maxsz + nfs4_label_maxsz + \
decode_mdsthreshold_maxsz))
#define nfs4_fattr_maxsz (nfs4_fattr_bitmap_maxsz + \
nfs4_owner_maxsz + \
nfs4_group_maxsz + \
nfs4_label_maxsz + \
- 4 + 4)
+ 1 + nfstime4_maxsz + \
+ 1 + nfstime4_maxsz)
#define encode_savefh_maxsz (op_encode_hdr_maxsz)
#define decode_savefh_maxsz (op_decode_hdr_maxsz)
#define encode_restorefh_maxsz (op_encode_hdr_maxsz)
WARN_ON_ONCE(xdr_stream_encode_u64(xdr, n) < 0);
}
+static ssize_t xdr_encode_bitmap4(struct xdr_stream *xdr,
+ const __u32 *bitmap, size_t len)
+{
+ ssize_t ret;
+
+ /* Trim empty words */
+ while (len > 0 && bitmap[len-1] == 0)
+ len--;
+ ret = xdr_stream_encode_uint32_array(xdr, bitmap, len);
+ if (WARN_ON_ONCE(ret < 0))
+ return ret;
+ return len;
+}
+
+static size_t mask_bitmap4(const __u32 *bitmap, const __u32 *mask,
+ __u32 *res, size_t len)
+{
+ size_t i;
+ __u32 tmp;
+
+ while (len > 0 && (bitmap[len-1] == 0 || mask[len-1] == 0))
+ len--;
+ for (i = len; i-- > 0;) {
+ tmp = bitmap[i] & mask[i];
+ res[i] = tmp;
+ }
+ return len;
+}
+
static void encode_nfs4_seqid(struct xdr_stream *xdr,
const struct nfs_seqid *seqid)
{
encode_opaque_fixed(xdr, verf->data, NFS4_VERIFIER_SIZE);
}
+static __be32 *
+xdr_encode_nfstime4(__be32 *p, const struct timespec *t)
+{
+ p = xdr_encode_hyper(p, (__s64)t->tv_sec);
+ *p++ = cpu_to_be32(t->tv_nsec);
+ return p;
+}
+
static void encode_attrs(struct xdr_stream *xdr, const struct iattr *iap,
const struct nfs4_label *label,
const umode_t *umask,
int owner_namelen = 0;
int owner_grouplen = 0;
__be32 *p;
- unsigned i;
uint32_t len = 0;
- uint32_t bmval_len;
uint32_t bmval[3] = { 0 };
/*
if (attrmask[1] & FATTR4_WORD1_TIME_ACCESS_SET) {
if (iap->ia_valid & ATTR_ATIME_SET) {
bmval[1] |= FATTR4_WORD1_TIME_ACCESS_SET;
- len += 16;
+ len += 4 + (nfstime4_maxsz << 2);
} else if (iap->ia_valid & ATTR_ATIME) {
bmval[1] |= FATTR4_WORD1_TIME_ACCESS_SET;
len += 4;
if (attrmask[1] & FATTR4_WORD1_TIME_MODIFY_SET) {
if (iap->ia_valid & ATTR_MTIME_SET) {
bmval[1] |= FATTR4_WORD1_TIME_MODIFY_SET;
- len += 16;
+ len += 4 + (nfstime4_maxsz << 2);
} else if (iap->ia_valid & ATTR_MTIME) {
bmval[1] |= FATTR4_WORD1_TIME_MODIFY_SET;
len += 4;
bmval[2] |= FATTR4_WORD2_SECURITY_LABEL;
}
- if (bmval[2] != 0)
- bmval_len = 3;
- else if (bmval[1] != 0)
- bmval_len = 2;
- else
- bmval_len = 1;
-
- p = reserve_space(xdr, 4 + (bmval_len << 2) + 4 + len);
-
- *p++ = cpu_to_be32(bmval_len);
- for (i = 0; i < bmval_len; i++)
- *p++ = cpu_to_be32(bmval[i]);
- *p++ = cpu_to_be32(len);
+ xdr_encode_bitmap4(xdr, bmval, ARRAY_SIZE(bmval));
+ xdr_stream_encode_opaque_inline(xdr, (void **)&p, len);
if (bmval[0] & FATTR4_WORD0_SIZE)
p = xdr_encode_hyper(p, iap->ia_size);
if (bmval[1] & FATTR4_WORD1_TIME_ACCESS_SET) {
if (iap->ia_valid & ATTR_ATIME_SET) {
*p++ = cpu_to_be32(NFS4_SET_TO_CLIENT_TIME);
- p = xdr_encode_hyper(p, (s64)iap->ia_atime.tv_sec);
- *p++ = cpu_to_be32(iap->ia_atime.tv_nsec);
+ p = xdr_encode_nfstime4(p, &iap->ia_atime);
} else
*p++ = cpu_to_be32(NFS4_SET_TO_SERVER_TIME);
}
if (bmval[1] & FATTR4_WORD1_TIME_MODIFY_SET) {
if (iap->ia_valid & ATTR_MTIME_SET) {
*p++ = cpu_to_be32(NFS4_SET_TO_CLIENT_TIME);
- p = xdr_encode_hyper(p, (s64)iap->ia_mtime.tv_sec);
- *p++ = cpu_to_be32(iap->ia_mtime.tv_nsec);
+ p = xdr_encode_nfstime4(p, &iap->ia_mtime);
} else
*p++ = cpu_to_be32(NFS4_SET_TO_SERVER_TIME);
}
create->server, create->server->attr_bitmask);
}
-static void encode_getattr_one(struct xdr_stream *xdr, uint32_t bitmap, struct compound_hdr *hdr)
-{
- __be32 *p;
-
- encode_op_hdr(xdr, OP_GETATTR, decode_getattr_maxsz, hdr);
- p = reserve_space(xdr, 8);
- *p++ = cpu_to_be32(1);
- *p = cpu_to_be32(bitmap);
-}
-
-static void encode_getattr_two(struct xdr_stream *xdr, uint32_t bm0, uint32_t bm1, struct compound_hdr *hdr)
-{
- __be32 *p;
-
- encode_op_hdr(xdr, OP_GETATTR, decode_getattr_maxsz, hdr);
- p = reserve_space(xdr, 12);
- *p++ = cpu_to_be32(2);
- *p++ = cpu_to_be32(bm0);
- *p = cpu_to_be32(bm1);
-}
-
-static void
-encode_getattr_three(struct xdr_stream *xdr,
- uint32_t bm0, uint32_t bm1, uint32_t bm2,
- struct compound_hdr *hdr)
+static void encode_getattr(struct xdr_stream *xdr,
+ const __u32 *bitmap, const __u32 *mask, size_t len,
+ struct compound_hdr *hdr)
{
- __be32 *p;
+ __u32 masked_bitmap[nfs4_fattr_bitmap_maxsz];
encode_op_hdr(xdr, OP_GETATTR, decode_getattr_maxsz, hdr);
- if (bm2) {
- p = reserve_space(xdr, 16);
- *p++ = cpu_to_be32(3);
- *p++ = cpu_to_be32(bm0);
- *p++ = cpu_to_be32(bm1);
- *p = cpu_to_be32(bm2);
- } else if (bm1) {
- p = reserve_space(xdr, 12);
- *p++ = cpu_to_be32(2);
- *p++ = cpu_to_be32(bm0);
- *p = cpu_to_be32(bm1);
- } else {
- p = reserve_space(xdr, 8);
- *p++ = cpu_to_be32(1);
- *p = cpu_to_be32(bm0);
+ if (mask) {
+ if (WARN_ON_ONCE(len > ARRAY_SIZE(masked_bitmap)))
+ len = ARRAY_SIZE(masked_bitmap);
+ len = mask_bitmap4(bitmap, mask, masked_bitmap, len);
+ bitmap = masked_bitmap;
}
+ xdr_encode_bitmap4(xdr, bitmap, len);
}
static void encode_getfattr(struct xdr_stream *xdr, const u32* bitmask, struct compound_hdr *hdr)
{
- encode_getattr_three(xdr, bitmask[0] & nfs4_fattr_bitmap[0],
- bitmask[1] & nfs4_fattr_bitmap[1],
- bitmask[2] & nfs4_fattr_bitmap[2],
- hdr);
+ encode_getattr(xdr, nfs4_fattr_bitmap, bitmask,
+ ARRAY_SIZE(nfs4_fattr_bitmap), hdr);
}
static void encode_getfattr_open(struct xdr_stream *xdr, const u32 *bitmask,
const u32 *open_bitmap,
struct compound_hdr *hdr)
{
- encode_getattr_three(xdr,
- bitmask[0] & open_bitmap[0],
- bitmask[1] & open_bitmap[1],
- bitmask[2] & open_bitmap[2],
- hdr);
+ encode_getattr(xdr, open_bitmap, bitmask, 3, hdr);
}
static void encode_fsinfo(struct xdr_stream *xdr, const u32* bitmask, struct compound_hdr *hdr)
{
- encode_getattr_three(xdr,
- bitmask[0] & nfs4_fsinfo_bitmap[0],
- bitmask[1] & nfs4_fsinfo_bitmap[1],
- bitmask[2] & nfs4_fsinfo_bitmap[2],
- hdr);
+ encode_getattr(xdr, nfs4_fsinfo_bitmap, bitmask,
+ ARRAY_SIZE(nfs4_fsinfo_bitmap), hdr);
}
static void encode_fs_locations(struct xdr_stream *xdr, const u32* bitmask, struct compound_hdr *hdr)
{
- encode_getattr_two(xdr, bitmask[0] & nfs4_fs_locations_bitmap[0],
- bitmask[1] & nfs4_fs_locations_bitmap[1], hdr);
+ encode_getattr(xdr, nfs4_fs_locations_bitmap, bitmask,
+ ARRAY_SIZE(nfs4_fs_locations_bitmap), hdr);
}
static void encode_getfh(struct xdr_stream *xdr, struct compound_hdr *hdr)
encode_sequence(xdr, &args->seq_args, &hdr);
encode_putfh(xdr, args->fh, &hdr);
encode_access(xdr, args->access, &hdr);
- encode_getfattr(xdr, args->bitmask, &hdr);
+ if (args->bitmask)
+ encode_getfattr(xdr, args->bitmask, &hdr);
encode_nops(&hdr);
}
struct compound_hdr hdr = {
.minorversion = nfs4_xdr_minorversion(&args->seq_args),
};
+ const __u32 nfs4_acl_bitmap[1] = {
+ [0] = FATTR4_WORD0_ACL,
+ };
uint32_t replen;
encode_compound_hdr(xdr, req, &hdr);
encode_sequence(xdr, &args->seq_args, &hdr);
encode_putfh(xdr, args->fh, &hdr);
replen = hdr.replen + op_decode_hdr_maxsz;
- encode_getattr_two(xdr, FATTR4_WORD0_ACL, 0, &hdr);
+ encode_getattr(xdr, nfs4_acl_bitmap, NULL,
+ ARRAY_SIZE(nfs4_acl_bitmap), &hdr);
xdr_inline_pages(&req->rq_rcv_buf, replen << 2,
args->acl_pages, 0, args->acl_len);
encode_compound_hdr(xdr, req, &hdr);
encode_sequence(xdr, &args->seq_args, &hdr);
encode_putfh(xdr, args->fh, &hdr);
- encode_getattr_one(xdr, args->bitmask[0] & nfs4_pathconf_bitmap[0],
- &hdr);
+ encode_getattr(xdr, nfs4_pathconf_bitmap, args->bitmask,
+ ARRAY_SIZE(nfs4_pathconf_bitmap), &hdr);
encode_nops(&hdr);
}
encode_compound_hdr(xdr, req, &hdr);
encode_sequence(xdr, &args->seq_args, &hdr);
encode_putfh(xdr, args->fh, &hdr);
- encode_getattr_two(xdr, args->bitmask[0] & nfs4_statfs_bitmap[0],
- args->bitmask[1] & nfs4_statfs_bitmap[1], &hdr);
+ encode_getattr(xdr, nfs4_statfs_bitmap, args->bitmask,
+ ARRAY_SIZE(nfs4_statfs_bitmap), &hdr);
encode_nops(&hdr);
}
encode_compound_hdr(xdr, req, &hdr);
encode_sequence(xdr, &args->seq_args, &hdr);
encode_putfh(xdr, args->fhandle, &hdr);
- encode_getattr_three(xdr, bitmask[0], bitmask[1], bitmask[2], &hdr);
+ encode_getattr(xdr, bitmask, NULL, 3, &hdr);
encode_nops(&hdr);
}
return -EIO;
}
-static int decode_attr_bitmap(struct xdr_stream *xdr, uint32_t *bitmap)
+static ssize_t
+decode_bitmap4(struct xdr_stream *xdr, uint32_t *bitmap, size_t sz)
{
- uint32_t bmlen;
- __be32 *p;
-
- p = xdr_inline_decode(xdr, 4);
- if (unlikely(!p))
- goto out_overflow;
- bmlen = be32_to_cpup(p);
+ ssize_t ret;
- bitmap[0] = bitmap[1] = bitmap[2] = 0;
- p = xdr_inline_decode(xdr, (bmlen << 2));
- if (unlikely(!p))
- goto out_overflow;
- if (bmlen > 0) {
- bitmap[0] = be32_to_cpup(p++);
- if (bmlen > 1) {
- bitmap[1] = be32_to_cpup(p++);
- if (bmlen > 2)
- bitmap[2] = be32_to_cpup(p);
- }
- }
- return 0;
-out_overflow:
+ ret = xdr_stream_decode_uint32_array(xdr, bitmap, sz);
+ if (likely(ret >= 0))
+ return ret;
+ if (ret == -EMSGSIZE)
+ return sz;
print_overflow_msg(__func__, xdr);
return -EIO;
}
+static int decode_attr_bitmap(struct xdr_stream *xdr, uint32_t *bitmap)
+{
+ ssize_t ret;
+ ret = decode_bitmap4(xdr, bitmap, 3);
+ return ret < 0 ? ret : 0;
+}
+
static int decode_attr_length(struct xdr_stream *xdr, uint32_t *attrlen, unsigned int *savep)
{
__be32 *p;
bitmap[1] &= ~FATTR4_WORD1_OWNER;
if (owner_name != NULL) {
- len = decode_nfs4_string(xdr, owner_name, GFP_NOWAIT);
+ len = decode_nfs4_string(xdr, owner_name, GFP_NOIO);
if (len <= 0)
goto out;
dprintk("%s: name=%s\n", __func__, owner_name->data);
bitmap[1] &= ~FATTR4_WORD1_OWNER_GROUP;
if (group_name != NULL) {
- len = decode_nfs4_string(xdr, group_name, GFP_NOWAIT);
+ len = decode_nfs4_string(xdr, group_name, GFP_NOIO);
if (len <= 0)
goto out;
dprintk("%s: name=%s\n", __func__, group_name->data);
return -EIO;
}
+static __be32 *
+xdr_decode_nfstime4(__be32 *p, struct timespec *t)
+{
+ __u64 sec;
+
+ p = xdr_decode_hyper(p, &sec);
+ t-> tv_sec = (time_t)sec;
+ t->tv_nsec = be32_to_cpup(p++);
+ return p;
+}
+
static int decode_attr_time(struct xdr_stream *xdr, struct timespec *time)
{
__be32 *p;
- uint64_t sec;
- uint32_t nsec;
- p = xdr_inline_decode(xdr, 12);
+ p = xdr_inline_decode(xdr, nfstime4_maxsz << 2);
if (unlikely(!p))
goto out_overflow;
- p = xdr_decode_hyper(p, &sec);
- nsec = be32_to_cpup(p);
- time->tv_sec = (time_t)sec;
- time->tv_nsec = (long)nsec;
+ xdr_decode_nfstime4(p, time);
return 0;
out_overflow:
print_overflow_msg(__func__, xdr);
static int decode_setattr(struct xdr_stream *xdr)
{
- __be32 *p;
- uint32_t bmlen;
int status;
status = decode_op_hdr(xdr, OP_SETATTR);
if (status)
return status;
- p = xdr_inline_decode(xdr, 4);
- if (unlikely(!p))
- goto out_overflow;
- bmlen = be32_to_cpup(p);
- p = xdr_inline_decode(xdr, bmlen << 2);
- if (likely(p))
+ if (decode_bitmap4(xdr, NULL, 0) >= 0)
return 0;
-out_overflow:
print_overflow_msg(__func__, xdr);
return -EIO;
}
status = decode_access(xdr, &res->supported, &res->access);
if (status != 0)
goto out;
- decode_getfattr(xdr, res->fattr, res->server);
+ if (res->fattr)
+ decode_getfattr(xdr, res->fattr, res->server);
out:
return status;
}
unsigned int savep;
uint32_t bitmap[3] = {0};
uint32_t len;
+ uint64_t new_cookie;
__be32 *p = xdr_inline_decode(xdr, 4);
if (unlikely(!p))
goto out_overflow;
p = xdr_inline_decode(xdr, 12);
if (unlikely(!p))
goto out_overflow;
- entry->prev_cookie = entry->cookie;
- p = xdr_decode_hyper(p, &entry->cookie);
+ p = xdr_decode_hyper(p, &new_cookie);
entry->len = be32_to_cpup(p);
p = xdr_inline_decode(xdr, entry->len);
if (entry->fattr->valid & NFS_ATTR_FATTR_TYPE)
entry->d_type = nfs_umode_to_dtype(entry->fattr->mode);
+ entry->prev_cookie = entry->cookie;
+ entry->cookie = new_cookie;
+
return 0;
out_overflow:
}
static int
-nfs_proc_remove(struct inode *dir, const struct qstr *name)
+nfs_proc_remove(struct inode *dir, struct dentry *dentry)
{
struct nfs_removeargs arg = {
.fh = NFS_FH(dir),
- .name = *name,
+ .name = dentry->d_name,
};
struct rpc_message msg = {
.rpc_proc = &nfs_procedures[NFSPROC_REMOVE],
};
int status;
- dprintk("NFS call remove %s\n", name->name);
+ dprintk("NFS call remove %pd2\n",dentry);
status = rpc_call_sync(NFS_CLIENT(dir), &msg, 0);
nfs_mark_for_revalidate(dir);
}
static void
-nfs_proc_unlink_setup(struct rpc_message *msg, struct inode *dir)
+nfs_proc_unlink_setup(struct rpc_message *msg, struct dentry *dentry)
{
msg->rpc_proc = &nfs_procedures[NFSPROC_REMOVE];
}
}
static void
-nfs_proc_rename_setup(struct rpc_message *msg, struct inode *dir)
+nfs_proc_rename_setup(struct rpc_message *msg,
+ struct dentry *old_dentry,
+ struct dentry *new_dentry)
{
msg->rpc_proc = &nfs_procedures[NFSPROC_RENAME];
}
return 0;
}
-static int nfs_return_delegation(struct inode *inode)
-{
- nfs_wb_all(inode);
- return 0;
-}
-
static const struct inode_operations nfs_dir_inode_operations = {
.create = nfs_create,
.lookup = nfs_lookup,
.lock_check_bounds = nfs_lock_check_bounds,
.close_context = nfs_close_context,
.have_delegation = nfs_have_delegation,
- .return_delegation = nfs_return_delegation,
.alloc_client = nfs_alloc_client,
.init_client = nfs_init_client,
.free_client = nfs_free_client,
data->args.fh = NFS_FH(dir);
nfs_fattr_init(data->res.dir_attr);
- NFS_PROTO(dir)->unlink_setup(&msg, dir);
+ NFS_PROTO(dir)->unlink_setup(&msg, data->dentry);
task_setup_data.rpc_client = NFS_CLIENT(dir);
task = rpc_run_task(&task_setup_data);
nfs_sb_active(old_dir->i_sb);
- NFS_PROTO(data->old_dir)->rename_setup(&msg, old_dir);
+ NFS_PROTO(data->old_dir)->rename_setup(&msg, old_dentry, new_dentry);
return rpc_run_task(&task_setup_data);
}
fileid = NFS_FILEID(d_inode(dentry));
- /* Return delegation in anticipation of the rename */
- NFS_PROTO(d_inode(dentry))->return_delegation(d_inode(dentry));
-
sdentry = NULL;
do {
int slen;
if (i_size >= end)
goto out;
i_size_write(inode, end);
+ NFS_I(inode)->cache_validity &= ~NFS_INO_INVALID_SIZE;
nfs_inc_stats(inode, NFSIOS_EXTENDWRITE);
out:
spin_unlock(&inode->i_lock);
}
/* Deal with the suid/sgid bit corner case */
- if (nfs_should_remove_suid(inode))
- nfs_mark_for_revalidate(inode);
+ if (nfs_should_remove_suid(inode)) {
+ spin_lock(&inode->i_lock);
+ NFS_I(inode)->cache_validity |= NFS_INO_INVALID_OTHER;
+ spin_unlock(&inode->i_lock);
+ }
return 0;
}
/*
* Cache validity bit flags
*/
-#define NFS_INO_INVALID_ATTR 0x0001 /* cached attrs are invalid */
-#define NFS_INO_INVALID_DATA 0x0002 /* cached data is invalid */
-#define NFS_INO_INVALID_ATIME 0x0004 /* cached atime is invalid */
-#define NFS_INO_INVALID_ACCESS 0x0008 /* cached access cred invalid */
-#define NFS_INO_INVALID_ACL 0x0010 /* cached acls are invalid */
-#define NFS_INO_REVAL_PAGECACHE 0x0020 /* must revalidate pagecache */
-#define NFS_INO_REVAL_FORCED 0x0040 /* force revalidation ignoring a delegation */
-#define NFS_INO_INVALID_LABEL 0x0080 /* cached label is invalid */
+#define NFS_INO_INVALID_DATA BIT(1) /* cached data is invalid */
+#define NFS_INO_INVALID_ATIME BIT(2) /* cached atime is invalid */
+#define NFS_INO_INVALID_ACCESS BIT(3) /* cached access cred invalid */
+#define NFS_INO_INVALID_ACL BIT(4) /* cached acls are invalid */
+#define NFS_INO_REVAL_PAGECACHE BIT(5) /* must revalidate pagecache */
+#define NFS_INO_REVAL_FORCED BIT(6) /* force revalidation ignoring a delegation */
+#define NFS_INO_INVALID_LABEL BIT(7) /* cached label is invalid */
+#define NFS_INO_INVALID_CHANGE BIT(8) /* cached change is invalid */
+#define NFS_INO_INVALID_CTIME BIT(9) /* cached ctime is invalid */
+#define NFS_INO_INVALID_MTIME BIT(10) /* cached mtime is invalid */
+#define NFS_INO_INVALID_SIZE BIT(11) /* cached size is invalid */
+#define NFS_INO_INVALID_OTHER BIT(12) /* other attrs are invalid */
+
+#define NFS_INO_INVALID_ATTR (NFS_INO_INVALID_CHANGE \
+ | NFS_INO_INVALID_CTIME \
+ | NFS_INO_INVALID_MTIME \
+ | NFS_INO_INVALID_SIZE \
+ | NFS_INO_INVALID_OTHER) /* inode metadata is invalid */
/*
* Bit offsets in flags field
struct nfs_inode *nfsi = NFS_I(inode);
spin_lock(&inode->i_lock);
- nfsi->cache_validity |= NFS_INO_INVALID_ATTR |
- NFS_INO_REVAL_PAGECACHE |
- NFS_INO_INVALID_ACCESS |
- NFS_INO_INVALID_ACL;
+ nfsi->cache_validity |= NFS_INO_REVAL_PAGECACHE
+ | NFS_INO_INVALID_ACCESS
+ | NFS_INO_INVALID_ACL
+ | NFS_INO_INVALID_CHANGE
+ | NFS_INO_INVALID_CTIME;
if (S_ISDIR(inode->i_mode))
nfsi->cache_validity |= NFS_INO_INVALID_DATA;
spin_unlock(&inode->i_lock);
unsigned int);
int (*create) (struct inode *, struct dentry *,
struct iattr *, int);
- int (*remove) (struct inode *, const struct qstr *);
- void (*unlink_setup) (struct rpc_message *, struct inode *dir);
+ int (*remove) (struct inode *, struct dentry *);
+ void (*unlink_setup) (struct rpc_message *, struct dentry *);
void (*unlink_rpc_prepare) (struct rpc_task *, struct nfs_unlinkdata *);
int (*unlink_done) (struct rpc_task *, struct inode *);
- void (*rename_setup) (struct rpc_message *msg, struct inode *dir);
+ void (*rename_setup) (struct rpc_message *msg,
+ struct dentry *old_dentry,
+ struct dentry *new_dentry);
void (*rename_rpc_prepare)(struct rpc_task *task, struct nfs_renamedata *);
int (*rename_done) (struct rpc_task *task, struct inode *old_dir, struct inode *new_dir);
int (*link) (struct inode *, struct inode *, const struct qstr *);
struct iattr *iattr,
int *);
int (*have_delegation)(struct inode *, fmode_t);
- int (*return_delegation)(struct inode *);
struct nfs_client *(*alloc_client) (const struct nfs_client_initdata *);
struct nfs_client *(*init_client) (struct nfs_client *,
const struct nfs_client_initdata *);
bool rpc_clnt_xprt_switch_has_addr(struct rpc_clnt *clnt,
const struct sockaddr *sap);
void rpc_cleanup_clids(void);
+
+static inline int rpc_reply_expected(struct rpc_task *task)
+{
+ return (task->tk_msg.rpc_proc != NULL) &&
+ (task->tk_msg.rpc_proc->p_decode != NULL);
+}
+
#endif /* __KERNEL__ */
#endif /* _LINUX_SUNRPC_CLNT_H */
return xdr->nwords << 2;
}
+ssize_t xdr_stream_decode_opaque(struct xdr_stream *xdr, void *ptr,
+ size_t size);
+ssize_t xdr_stream_decode_opaque_dup(struct xdr_stream *xdr, void **ptr,
+ size_t maxlen, gfp_t gfp_flags);
+ssize_t xdr_stream_decode_string(struct xdr_stream *xdr, char *str,
+ size_t size);
ssize_t xdr_stream_decode_string_dup(struct xdr_stream *xdr, char **str,
size_t maxlen, gfp_t gfp_flags);
/**
return len;
}
+/**
+ * xdr_stream_encode_opaque_inline - Encode opaque xdr data
+ * @xdr: pointer to xdr_stream
+ * @ptr: pointer to void pointer
+ * @len: size of object
+ *
+ * Return values:
+ * On success, returns length in bytes of XDR buffer consumed
+ * %-EMSGSIZE on XDR buffer overflow
+ */
+static inline ssize_t
+xdr_stream_encode_opaque_inline(struct xdr_stream *xdr, void **ptr, size_t len)
+{
+ size_t count = sizeof(__u32) + xdr_align_size(len);
+ __be32 *p = xdr_reserve_space(xdr, count);
+
+ if (unlikely(!p)) {
+ *ptr = NULL;
+ return -EMSGSIZE;
+ }
+ xdr_encode_opaque(p, NULL, len);
+ *ptr = ++p;
+ return count;
+}
+
/**
* xdr_stream_encode_opaque_fixed - Encode fixed length opaque xdr data
* @xdr: pointer to xdr_stream
return count;
}
+/**
+ * xdr_stream_encode_uint32_array - Encode variable length array of integers
+ * @xdr: pointer to xdr_stream
+ * @array: array of integers
+ * @array_size: number of elements in @array
+ *
+ * Return values:
+ * On success, returns length in bytes of XDR buffer consumed
+ * %-EMSGSIZE on XDR buffer overflow
+ */
+static inline ssize_t
+xdr_stream_encode_uint32_array(struct xdr_stream *xdr,
+ const __u32 *array, size_t array_size)
+{
+ ssize_t ret = (array_size+1) * sizeof(__u32);
+ __be32 *p = xdr_reserve_space(xdr, ret);
+
+ if (unlikely(!p))
+ return -EMSGSIZE;
+ *p++ = cpu_to_be32(array_size);
+ for (; array_size > 0; p++, array++, array_size--)
+ *p = cpu_to_be32p(array);
+ return ret;
+}
+
/**
* xdr_stream_decode_u32 - Decode a 32-bit integer
* @xdr: pointer to xdr_stream
}
return len;
}
+
+/**
+ * xdr_stream_decode_uint32_array - Decode variable length array of integers
+ * @xdr: pointer to xdr_stream
+ * @array: location to store the integer array or NULL
+ * @array_size: number of elements to store
+ *
+ * Return values:
+ * On success, returns number of elements stored in @array
+ * %-EBADMSG on XDR buffer overflow
+ * %-EMSGSIZE if the size of the array exceeds @array_size
+ */
+static inline ssize_t
+xdr_stream_decode_uint32_array(struct xdr_stream *xdr,
+ __u32 *array, size_t array_size)
+{
+ __be32 *p;
+ __u32 len;
+ ssize_t retval;
+
+ if (unlikely(xdr_stream_decode_u32(xdr, &len) < 0))
+ return -EBADMSG;
+ p = xdr_inline_decode(xdr, len * sizeof(*p));
+ if (unlikely(!p))
+ return -EBADMSG;
+ if (array == NULL)
+ return len;
+ if (len <= array_size) {
+ if (len < array_size)
+ memset(array+len, 0, (array_size-len)*sizeof(*array));
+ array_size = len;
+ retval = len;
+ } else
+ retval = -EMSGSIZE;
+ for (; array_size > 0; p++, array++, array_size--)
+ *array = be32_to_cpup(p);
+ return retval;
+}
#endif /* __KERNEL__ */
#endif /* _SUNRPC_XDR_H_ */
struct list_head free; /* free slots */
unsigned int max_reqs; /* max number of slots */
unsigned int min_reqs; /* min number of slots */
- atomic_t num_reqs; /* total slots */
+ unsigned int num_reqs; /* total slots */
unsigned long state; /* transport state */
unsigned char resvport : 1; /* use a reserved port */
atomic_t swapper; /* we're swapping over this
void xprt_write_space(struct rpc_xprt *xprt);
void xprt_adjust_cwnd(struct rpc_xprt *xprt, struct rpc_task *task, int result);
struct rpc_rqst * xprt_lookup_rqst(struct rpc_xprt *xprt, __be32 xid);
+void xprt_update_rtt(struct rpc_task *task);
void xprt_complete_rqst(struct rpc_task *task, int copied);
void xprt_pin_rqst(struct rpc_rqst *req);
void xprt_unpin_rqst(struct rpc_rqst *req);
);
TRACE_EVENT(rpc_connect_status,
- TP_PROTO(struct rpc_task *task, int status),
+ TP_PROTO(const struct rpc_task *task),
- TP_ARGS(task, status),
+ TP_ARGS(task),
TP_STRUCT__entry(
__field(unsigned int, task_id)
TP_fast_assign(
__entry->task_id = task->tk_pid;
__entry->client_id = task->tk_client->cl_clid;
- __entry->status = status;
+ __entry->status = task->tk_status;
),
TP_printk("task:%u@%u status=%d",
DECLARE_EVENT_CLASS(rpc_task_running,
- TP_PROTO(const struct rpc_clnt *clnt, const struct rpc_task *task, const void *action),
+ TP_PROTO(const struct rpc_task *task, const void *action),
- TP_ARGS(clnt, task, action),
+ TP_ARGS(task, action),
TP_STRUCT__entry(
__field(unsigned int, task_id)
),
TP_fast_assign(
- __entry->client_id = clnt ? clnt->cl_clid : -1;
+ __entry->client_id = task->tk_client ?
+ task->tk_client->cl_clid : -1;
__entry->task_id = task->tk_pid;
__entry->action = action;
__entry->runstate = task->tk_runstate;
DEFINE_EVENT(rpc_task_running, rpc_task_begin,
- TP_PROTO(const struct rpc_clnt *clnt, const struct rpc_task *task, const void *action),
+ TP_PROTO(const struct rpc_task *task, const void *action),
- TP_ARGS(clnt, task, action)
+ TP_ARGS(task, action)
);
DEFINE_EVENT(rpc_task_running, rpc_task_run_action,
- TP_PROTO(const struct rpc_clnt *clnt, const struct rpc_task *task, const void *action),
+ TP_PROTO(const struct rpc_task *task, const void *action),
- TP_ARGS(clnt, task, action)
+ TP_ARGS(task, action)
);
DEFINE_EVENT(rpc_task_running, rpc_task_complete,
- TP_PROTO(const struct rpc_clnt *clnt, const struct rpc_task *task, const void *action),
+ TP_PROTO(const struct rpc_task *task, const void *action),
- TP_ARGS(clnt, task, action)
+ TP_ARGS(task, action)
);
DECLARE_EVENT_CLASS(rpc_task_queued,
- TP_PROTO(const struct rpc_clnt *clnt, const struct rpc_task *task, const struct rpc_wait_queue *q),
+ TP_PROTO(const struct rpc_task *task, const struct rpc_wait_queue *q),
- TP_ARGS(clnt, task, q),
+ TP_ARGS(task, q),
TP_STRUCT__entry(
__field(unsigned int, task_id)
),
TP_fast_assign(
- __entry->client_id = clnt ? clnt->cl_clid : -1;
+ __entry->client_id = task->tk_client ?
+ task->tk_client->cl_clid : -1;
__entry->task_id = task->tk_pid;
__entry->timeout = task->tk_timeout;
__entry->runstate = task->tk_runstate;
DEFINE_EVENT(rpc_task_queued, rpc_task_sleep,
- TP_PROTO(const struct rpc_clnt *clnt, const struct rpc_task *task, const struct rpc_wait_queue *q),
+ TP_PROTO(const struct rpc_task *task, const struct rpc_wait_queue *q),
- TP_ARGS(clnt, task, q)
+ TP_ARGS(task, q)
);
DEFINE_EVENT(rpc_task_queued, rpc_task_wakeup,
- TP_PROTO(const struct rpc_clnt *clnt, const struct rpc_task *task, const struct rpc_wait_queue *q),
+ TP_PROTO(const struct rpc_task *task, const struct rpc_wait_queue *q),
+
+ TP_ARGS(task, q)
+
+);
+
+TRACE_EVENT(rpc_stats_latency,
+
+ TP_PROTO(
+ const struct rpc_task *task,
+ ktime_t backlog,
+ ktime_t rtt,
+ ktime_t execute
+ ),
- TP_ARGS(clnt, task, q)
+ TP_ARGS(task, backlog, rtt, execute),
+ TP_STRUCT__entry(
+ __field(u32, xid)
+ __field(int, version)
+ __string(progname, task->tk_client->cl_program->name)
+ __string(procname, rpc_proc_name(task))
+ __field(unsigned long, backlog)
+ __field(unsigned long, rtt)
+ __field(unsigned long, execute)
+ __string(addr,
+ task->tk_xprt->address_strings[RPC_DISPLAY_ADDR])
+ __string(port,
+ task->tk_xprt->address_strings[RPC_DISPLAY_PORT])
+ ),
+
+ TP_fast_assign(
+ __entry->xid = be32_to_cpu(task->tk_rqstp->rq_xid);
+ __entry->version = task->tk_client->cl_vers;
+ __assign_str(progname, task->tk_client->cl_program->name)
+ __assign_str(procname, rpc_proc_name(task))
+ __entry->backlog = ktime_to_us(backlog);
+ __entry->rtt = ktime_to_us(rtt);
+ __entry->execute = ktime_to_us(execute);
+ __assign_str(addr,
+ task->tk_xprt->address_strings[RPC_DISPLAY_ADDR]);
+ __assign_str(port,
+ task->tk_xprt->address_strings[RPC_DISPLAY_PORT]);
+ ),
+
+ TP_printk("peer=[%s]:%s xid=0x%08x %sv%d %s backlog=%lu rtt=%lu execute=%lu",
+ __get_str(addr), __get_str(port), __entry->xid,
+ __get_str(progname), __entry->version, __get_str(procname),
+ __entry->backlog, __entry->rtt, __entry->execute)
);
/*
TP_PROTO(struct rpc_xprt *xprt, __be32 xid, int status),
TP_ARGS(xprt, xid, status));
+TRACE_EVENT(xprt_ping,
+ TP_PROTO(const struct rpc_xprt *xprt, int status),
+
+ TP_ARGS(xprt, status),
+
+ TP_STRUCT__entry(
+ __field(int, status)
+ __string(addr, xprt->address_strings[RPC_DISPLAY_ADDR])
+ __string(port, xprt->address_strings[RPC_DISPLAY_PORT])
+ ),
+
+ TP_fast_assign(
+ __entry->status = status;
+ __assign_str(addr, xprt->address_strings[RPC_DISPLAY_ADDR]);
+ __assign_str(port, xprt->address_strings[RPC_DISPLAY_PORT]);
+ ),
+
+ TP_printk("peer=[%s]:%s status=%d",
+ __get_str(addr), __get_str(port), __entry->status)
+);
+
TRACE_EVENT(xs_tcp_data_ready,
TP_PROTO(struct rpc_xprt *xprt, int err, unsigned int total),
dprint_status(task);
- trace_rpc_connect_status(task, status);
+ trace_rpc_connect_status(task);
task->tk_status = 0;
switch (status) {
case -ECONNREFUSED:
case -EPERM:
if (RPC_IS_SOFTCONN(task)) {
xprt_end_transmit(task);
+ if (!task->tk_msg.rpc_proc->p_proc)
+ trace_xprt_ping(task->tk_xprt,
+ task->tk_status);
rpc_exit(task, task->tk_status);
break;
}
struct rpc_rqst *req = task->tk_rqstp;
int status;
+ if (!task->tk_msg.rpc_proc->p_proc)
+ trace_xprt_ping(task->tk_xprt, task->tk_status);
+
if (req->rq_reply_bytes_recvd > 0 && !req->rq_bytes_sent)
task->tk_status = req->rq_reply_bytes_recvd;
{
rpc_task_set_debuginfo(task);
set_bit(RPC_TASK_ACTIVE, &task->tk_runstate);
- trace_rpc_task_begin(task->tk_client, task, NULL);
+ trace_rpc_task_begin(task, NULL);
}
/*
unsigned long flags;
int ret;
- trace_rpc_task_complete(task->tk_client, task, NULL);
+ trace_rpc_task_complete(task, NULL);
spin_lock_irqsave(&wq->lock, flags);
clear_bit(RPC_TASK_ACTIVE, &task->tk_runstate);
dprintk("RPC: %5u sleep_on(queue \"%s\" time %lu)\n",
task->tk_pid, rpc_qname(q), jiffies);
- trace_rpc_task_sleep(task->tk_client, task, q);
+ trace_rpc_task_sleep(task, q);
__rpc_add_wait_queue(q, task, queue_priority);
return;
}
- trace_rpc_task_wakeup(task->tk_client, task, queue);
+ trace_rpc_task_wakeup(task, queue);
__rpc_remove_wait_queue(queue, task);
}
if (!do_action)
break;
- trace_rpc_task_run_action(task->tk_client, task, do_action);
+ trace_rpc_task_run_action(task, do_action);
do_action(task);
/*
#include <linux/sunrpc/metrics.h>
#include <linux/rcupdate.h>
+#include <trace/events/sunrpc.h>
+
#include "netns.h"
#define RPCDBG_FACILITY RPCDBG_MISC
struct rpc_iostats *op_metrics)
{
struct rpc_rqst *req = task->tk_rqstp;
- ktime_t delta, now;
+ ktime_t backlog, execute, now;
if (!op_metrics || !req)
return;
op_metrics->om_bytes_sent += req->rq_xmit_bytes_sent;
op_metrics->om_bytes_recv += req->rq_reply_bytes_recvd;
+ backlog = 0;
if (ktime_to_ns(req->rq_xtime)) {
- delta = ktime_sub(req->rq_xtime, task->tk_start);
- op_metrics->om_queue = ktime_add(op_metrics->om_queue, delta);
+ backlog = ktime_sub(req->rq_xtime, task->tk_start);
+ op_metrics->om_queue = ktime_add(op_metrics->om_queue, backlog);
}
+
op_metrics->om_rtt = ktime_add(op_metrics->om_rtt, req->rq_rtt);
- delta = ktime_sub(now, task->tk_start);
- op_metrics->om_execute = ktime_add(op_metrics->om_execute, delta);
+ execute = ktime_sub(now, task->tk_start);
+ op_metrics->om_execute = ktime_add(op_metrics->om_execute, execute);
spin_unlock(&op_metrics->om_lock);
+
+ trace_rpc_stats_latency(req->rq_task, backlog, req->rq_rtt, execute);
}
EXPORT_SYMBOL_GPL(rpc_count_iostats_metrics);
char data[];
};
-static inline int rpc_reply_expected(struct rpc_task *task)
-{
- return (task->tk_msg.rpc_proc != NULL) &&
- (task->tk_msg.rpc_proc->p_decode != NULL);
-}
-
static inline int sock_is_loopback(struct sock *sk)
{
struct dst_entry *dst;
}
EXPORT_SYMBOL_GPL(xdr_process_buf);
+/**
+ * xdr_stream_decode_opaque - Decode variable length opaque
+ * @xdr: pointer to xdr_stream
+ * @ptr: location to store opaque data
+ * @size: size of storage buffer @ptr
+ *
+ * Return values:
+ * On success, returns size of object stored in *@ptr
+ * %-EBADMSG on XDR buffer overflow
+ * %-EMSGSIZE on overflow of storage buffer @ptr
+ */
+ssize_t xdr_stream_decode_opaque(struct xdr_stream *xdr, void *ptr, size_t size)
+{
+ ssize_t ret;
+ void *p;
+
+ ret = xdr_stream_decode_opaque_inline(xdr, &p, size);
+ if (ret <= 0)
+ return ret;
+ memcpy(ptr, p, ret);
+ return ret;
+}
+EXPORT_SYMBOL_GPL(xdr_stream_decode_opaque);
+
+/**
+ * xdr_stream_decode_opaque_dup - Decode and duplicate variable length opaque
+ * @xdr: pointer to xdr_stream
+ * @ptr: location to store pointer to opaque data
+ * @maxlen: maximum acceptable object size
+ * @gfp_flags: GFP mask to use
+ *
+ * Return values:
+ * On success, returns size of object stored in *@ptr
+ * %-EBADMSG on XDR buffer overflow
+ * %-EMSGSIZE if the size of the object would exceed @maxlen
+ * %-ENOMEM on memory allocation failure
+ */
+ssize_t xdr_stream_decode_opaque_dup(struct xdr_stream *xdr, void **ptr,
+ size_t maxlen, gfp_t gfp_flags)
+{
+ ssize_t ret;
+ void *p;
+
+ ret = xdr_stream_decode_opaque_inline(xdr, &p, maxlen);
+ if (ret > 0) {
+ *ptr = kmemdup(p, ret, gfp_flags);
+ if (*ptr != NULL)
+ return ret;
+ ret = -ENOMEM;
+ }
+ *ptr = NULL;
+ return ret;
+}
+EXPORT_SYMBOL_GPL(xdr_stream_decode_opaque_dup);
+
+/**
+ * xdr_stream_decode_string - Decode variable length string
+ * @xdr: pointer to xdr_stream
+ * @str: location to store string
+ * @size: size of storage buffer @str
+ *
+ * Return values:
+ * On success, returns length of NUL-terminated string stored in *@str
+ * %-EBADMSG on XDR buffer overflow
+ * %-EMSGSIZE on overflow of storage buffer @str
+ */
+ssize_t xdr_stream_decode_string(struct xdr_stream *xdr, char *str, size_t size)
+{
+ ssize_t ret;
+ void *p;
+
+ ret = xdr_stream_decode_opaque_inline(xdr, &p, size);
+ if (ret > 0) {
+ memcpy(str, p, ret);
+ str[ret] = '\0';
+ return strlen(str);
+ }
+ *str = '\0';
+ return ret;
+}
+EXPORT_SYMBOL_GPL(xdr_stream_decode_string);
+
/**
* xdr_stream_decode_string_dup - Decode and duplicate variable length string
* @xdr: pointer to xdr_stream
* @xprt: transport on which the original request was transmitted
* @xid: RPC XID of incoming reply
*
+ * Caller holds xprt->recv_lock.
*/
struct rpc_rqst *xprt_lookup_rqst(struct rpc_xprt *xprt, __be32 xid)
{
list_for_each_entry(entry, &xprt->recv, rq_list)
if (entry->rq_xid == xid) {
trace_xprt_lookup_rqst(xprt, xid, 0);
+ entry->rq_rtt = ktime_sub(ktime_get(), entry->rq_xtime);
return entry;
}
}
}
-static void xprt_update_rtt(struct rpc_task *task)
+/**
+ * xprt_update_rtt - Update RPC RTT statistics
+ * @task: RPC request that recently completed
+ *
+ * Caller holds xprt->recv_lock.
+ */
+void xprt_update_rtt(struct rpc_task *task)
{
struct rpc_rqst *req = task->tk_rqstp;
struct rpc_rtt *rtt = task->tk_client->cl_rtt;
rpc_set_timeo(rtt, timer, req->rq_ntrans - 1);
}
}
+EXPORT_SYMBOL_GPL(xprt_update_rtt);
/**
* xprt_complete_rqst - called when reply processing is complete
* @task: RPC request that recently completed
* @copied: actual number of bytes received from the transport
*
- * Caller holds transport lock.
+ * Caller holds xprt->recv_lock.
*/
void xprt_complete_rqst(struct rpc_task *task, int copied)
{
trace_xprt_complete_rqst(xprt, req->rq_xid, copied);
xprt->stat.recvs++;
- req->rq_rtt = ktime_sub(ktime_get(), req->rq_xtime);
- if (xprt->ops->timer != NULL)
- xprt_update_rtt(task);
list_del_init(&req->rq_list);
req->rq_private_buf.len = copied;
struct rpc_rqst *req = task->tk_rqstp;
struct rpc_xprt *xprt = req->rq_xprt;
unsigned int connect_cookie;
- int status, numreqs;
+ int status;
dprintk("RPC: %5u xprt_transmit(%u)\n", task->tk_pid, req->rq_slen);
return;
connect_cookie = xprt->connect_cookie;
- req->rq_xtime = ktime_get();
status = xprt->ops->send_request(task);
trace_xprt_transmit(xprt, req->rq_xid, status);
if (status != 0) {
xprt->ops->set_retrans_timeout(task);
- numreqs = atomic_read(&xprt->num_reqs);
- if (numreqs > xprt->stat.max_slots)
- xprt->stat.max_slots = numreqs;
xprt->stat.sends++;
xprt->stat.req_u += xprt->stat.sends - xprt->stat.recvs;
xprt->stat.bklog_u += xprt->backlog.qlen;
{
struct rpc_rqst *req = ERR_PTR(-EAGAIN);
- if (!atomic_add_unless(&xprt->num_reqs, 1, xprt->max_reqs))
+ if (xprt->num_reqs >= xprt->max_reqs)
goto out;
+ ++xprt->num_reqs;
spin_unlock(&xprt->reserve_lock);
req = kzalloc(sizeof(struct rpc_rqst), GFP_NOFS);
spin_lock(&xprt->reserve_lock);
if (req != NULL)
goto out;
- atomic_dec(&xprt->num_reqs);
+ --xprt->num_reqs;
req = ERR_PTR(-ENOMEM);
out:
return req;
static bool xprt_dynamic_free_slot(struct rpc_xprt *xprt, struct rpc_rqst *req)
{
- if (atomic_add_unless(&xprt->num_reqs, -1, xprt->min_reqs)) {
+ if (xprt->num_reqs > xprt->min_reqs) {
+ --xprt->num_reqs;
kfree(req);
return true;
}
spin_unlock(&xprt->reserve_lock);
return;
out_init_req:
+ xprt->stat.max_slots = max_t(unsigned int, xprt->stat.max_slots,
+ xprt->num_reqs);
task->tk_status = 0;
task->tk_rqstp = req;
xprt_request_init(task, xprt);
else
xprt->max_reqs = num_prealloc;
xprt->min_reqs = num_prealloc;
- atomic_set(&xprt->num_reqs, num_prealloc);
+ xprt->num_reqs = num_prealloc;
return xprt;
if (IS_ERR(req))
return PTR_ERR(req);
- rb = rpcrdma_alloc_regbuf(RPCRDMA_HDRBUF_SIZE,
- DMA_TO_DEVICE, GFP_KERNEL);
- if (IS_ERR(rb))
- goto out_fail;
- req->rl_rdmabuf = rb;
- xdr_buf_init(&req->rl_hdrbuf, rb->rg_base, rdmab_length(rb));
-
size = r_xprt->rx_data.inline_rsize;
rb = rpcrdma_alloc_regbuf(size, DMA_TO_DEVICE, GFP_KERNEL);
if (IS_ERR(rb))
mr = rpcrdma_mr_get(r_xprt);
if (!mr)
- return ERR_PTR(-ENOBUFS);
+ return ERR_PTR(-EAGAIN);
pageoff = offset_in_page(seg1->mr_offset);
seg1->mr_offset -= pageoff; /* start of page */
return ERR_PTR(-EIO);
}
+/* Post Send WR containing the RPC Call message.
+ */
+static int
+fmr_op_send(struct rpcrdma_ia *ia, struct rpcrdma_req *req)
+{
+ struct ib_send_wr *bad_wr;
+
+ return ib_post_send(ia->ri_id->qp, &req->rl_sendctx->sc_wr, &bad_wr);
+}
+
/* Invalidate all memory regions that were registered for "req".
*
* Sleeps until it is safe for the host CPU to access the
const struct rpcrdma_memreg_ops rpcrdma_fmr_memreg_ops = {
.ro_map = fmr_op_map,
+ .ro_send = fmr_op_send,
.ro_unmap_sync = fmr_op_unmap_sync,
.ro_recover_mr = fmr_op_recover_mr,
.ro_open = fmr_op_open,
struct rpcrdma_mr *mr;
struct ib_mr *ibmr;
struct ib_reg_wr *reg_wr;
- struct ib_send_wr *bad_wr;
- int rc, i, n;
+ int i, n;
u8 key;
mr = NULL;
rpcrdma_mr_defer_recovery(mr);
mr = rpcrdma_mr_get(r_xprt);
if (!mr)
- return ERR_PTR(-ENOBUFS);
+ return ERR_PTR(-EAGAIN);
} while (mr->frwr.fr_state != FRWR_IS_INVALID);
frwr = &mr->frwr;
frwr->fr_state = FRWR_IS_VALID;
ib_update_fast_reg_key(ibmr, ++key);
reg_wr = &frwr->fr_regwr;
- reg_wr->wr.next = NULL;
- reg_wr->wr.opcode = IB_WR_REG_MR;
- frwr->fr_cqe.done = frwr_wc_fastreg;
- reg_wr->wr.wr_cqe = &frwr->fr_cqe;
- reg_wr->wr.num_sge = 0;
- reg_wr->wr.send_flags = 0;
reg_wr->mr = ibmr;
reg_wr->key = ibmr->rkey;
reg_wr->access = writing ?
IB_ACCESS_REMOTE_WRITE | IB_ACCESS_LOCAL_WRITE :
IB_ACCESS_REMOTE_READ;
- rc = ib_post_send(ia->ri_id->qp, ®_wr->wr, &bad_wr);
- if (rc)
- goto out_senderr;
-
mr->mr_handle = ibmr->rkey;
mr->mr_length = ibmr->length;
mr->mr_offset = ibmr->iova;
frwr->fr_mr, n, mr->mr_nents);
rpcrdma_mr_defer_recovery(mr);
return ERR_PTR(-EIO);
+}
-out_senderr:
- pr_err("rpcrdma: FRWR registration ib_post_send returned %i\n", rc);
- rpcrdma_mr_defer_recovery(mr);
- return ERR_PTR(-ENOTCONN);
+/* Post Send WR containing the RPC Call message.
+ *
+ * For FRMR, chain any FastReg WRs to the Send WR. Only a
+ * single ib_post_send call is needed to register memory
+ * and then post the Send WR.
+ */
+static int
+frwr_op_send(struct rpcrdma_ia *ia, struct rpcrdma_req *req)
+{
+ struct ib_send_wr *post_wr, *bad_wr;
+ struct rpcrdma_mr *mr;
+
+ post_wr = &req->rl_sendctx->sc_wr;
+ list_for_each_entry(mr, &req->rl_registered, mr_list) {
+ struct rpcrdma_frwr *frwr;
+
+ frwr = &mr->frwr;
+
+ frwr->fr_cqe.done = frwr_wc_fastreg;
+ frwr->fr_regwr.wr.next = post_wr;
+ frwr->fr_regwr.wr.wr_cqe = &frwr->fr_cqe;
+ frwr->fr_regwr.wr.num_sge = 0;
+ frwr->fr_regwr.wr.opcode = IB_WR_REG_MR;
+ frwr->fr_regwr.wr.send_flags = 0;
+
+ post_wr = &frwr->fr_regwr.wr;
+ }
+
+ /* If ib_post_send fails, the next ->send_request for
+ * @req will queue these MWs for recovery.
+ */
+ return ib_post_send(ia->ri_id->qp, post_wr, &bad_wr);
}
/* Handle a remotely invalidated mr on the @mrs list
const struct rpcrdma_memreg_ops rpcrdma_frwr_memreg_ops = {
.ro_map = frwr_op_map,
+ .ro_send = frwr_op_send,
.ro_reminv = frwr_op_reminv,
.ro_unmap_sync = frwr_op_unmap_sync,
.ro_recover_mr = frwr_op_recover_mr,
seg = r_xprt->rx_ia.ri_ops->ro_map(r_xprt, seg, nsegs,
false, &mr);
if (IS_ERR(seg))
- return PTR_ERR(seg);
+ goto out_maperr;
rpcrdma_mr_push(mr, &req->rl_registered);
if (encode_read_segment(xdr, mr, pos) < 0)
} while (nsegs);
return 0;
+
+out_maperr:
+ if (PTR_ERR(seg) == -EAGAIN)
+ xprt_wait_for_buffer_space(rqst->rq_task, NULL);
+ return PTR_ERR(seg);
}
/* Register and XDR encode the Write list. Supports encoding a list
seg = r_xprt->rx_ia.ri_ops->ro_map(r_xprt, seg, nsegs,
true, &mr);
if (IS_ERR(seg))
- return PTR_ERR(seg);
+ goto out_maperr;
rpcrdma_mr_push(mr, &req->rl_registered);
if (encode_rdma_segment(xdr, mr) < 0)
*segcount = cpu_to_be32(nchunks);
return 0;
+
+out_maperr:
+ if (PTR_ERR(seg) == -EAGAIN)
+ xprt_wait_for_buffer_space(rqst->rq_task, NULL);
+ return PTR_ERR(seg);
}
/* Register and XDR encode the Reply chunk. Supports encoding an array
seg = r_xprt->rx_ia.ri_ops->ro_map(r_xprt, seg, nsegs,
true, &mr);
if (IS_ERR(seg))
- return PTR_ERR(seg);
+ goto out_maperr;
rpcrdma_mr_push(mr, &req->rl_registered);
if (encode_rdma_segment(xdr, mr) < 0)
*segcount = cpu_to_be32(nchunks);
return 0;
+
+out_maperr:
+ if (PTR_ERR(seg) == -EAGAIN)
+ xprt_wait_for_buffer_space(rqst->rq_task, NULL);
+ return PTR_ERR(seg);
}
/**
* Returns:
* %0 if the RPC was sent successfully,
* %-ENOTCONN if the connection was lost,
- * %-EAGAIN if not enough pages are available for on-demand reply buffer,
- * %-ENOBUFS if no MRs are available to register chunks,
+ * %-EAGAIN if the caller should call again with the same arguments,
+ * %-ENOBUFS if the caller should call again after a delay,
* %-EMSGSIZE if the transport header is too small,
* %-EIO if a permanent problem occurred while marshaling.
*/
return 0;
out_err:
- if (ret != -ENOBUFS) {
- pr_err("rpcrdma: header marshaling failed (%d)\n", ret);
- r_xprt->rx_stats.failed_marshal_count++;
- }
+ r_xprt->rx_stats.failed_marshal_count++;
return ret;
}
trace_xprtrdma_reply(rqst->rq_task, rep, req, credits);
- queue_work_on(req->rl_cpu, rpcrdma_receive_wq, &rep->rr_work);
+ queue_work(rpcrdma_receive_wq, &rep->rr_work);
return;
out_badstatus:
#include <linux/slab.h>
#include <linux/seq_file.h>
#include <linux/sunrpc/addr.h>
-#include <linux/smp.h>
#include "xprt_rdma.h"
struct rpc_xprt *xprt = &r_xprt->rx_xprt;
spin_lock_bh(&xprt->transport_lock);
- if (++xprt->connect_cookie == 0) /* maintain a reserved value */
- ++xprt->connect_cookie;
if (ep->rep_connected > 0) {
if (!xprt_test_and_set_connected(xprt))
xprt_wake_pending_tasks(xprt, 0);
}
}
-/* Allocate a fixed-size buffer in which to construct and send the
- * RPC-over-RDMA header for this request.
- */
-static bool
-rpcrdma_get_rdmabuf(struct rpcrdma_xprt *r_xprt, struct rpcrdma_req *req,
- gfp_t flags)
-{
- size_t size = RPCRDMA_HDRBUF_SIZE;
- struct rpcrdma_regbuf *rb;
-
- if (req->rl_rdmabuf)
- return true;
-
- rb = rpcrdma_alloc_regbuf(size, DMA_TO_DEVICE, flags);
- if (IS_ERR(rb))
- return false;
-
- r_xprt->rx_stats.hardway_register_count += size;
- req->rl_rdmabuf = rb;
- xdr_buf_init(&req->rl_hdrbuf, rb->rg_base, rdmab_length(rb));
- return true;
-}
-
static bool
rpcrdma_get_sendbuf(struct rpcrdma_xprt *r_xprt, struct rpcrdma_req *req,
size_t size, gfp_t flags)
if (RPC_IS_SWAPPER(task))
flags = __GFP_MEMALLOC | GFP_NOWAIT | __GFP_NOWARN;
- if (!rpcrdma_get_rdmabuf(r_xprt, req, flags))
- goto out_fail;
if (!rpcrdma_get_sendbuf(r_xprt, req, rqst->rq_callsize, flags))
goto out_fail;
if (!rpcrdma_get_recvbuf(r_xprt, req, rqst->rq_rcvsize, flags))
goto out_fail;
- req->rl_cpu = smp_processor_id();
- req->rl_connect_cookie = 0; /* our reserved value */
rpcrdma_set_xprtdata(rqst, req);
rqst->rq_buffer = req->rl_sendbuf->rg_base;
rqst->rq_rbuffer = req->rl_recvbuf->rg_base;
* Returns:
* %0 if the RPC message has been sent
* %-ENOTCONN if the caller should reconnect and call again
- * %-ENOBUFS if the caller should call again later
+ * %-EAGAIN if the caller should call again
+ * %-ENOBUFS if the caller should call again after a delay
* %-EIO if a permanent error occurred and the request was not
* sent. Do not try to send this message again.
*/
rpcrdma_recv_buffer_get(req);
/* Must suppress retransmit to maintain credits */
- if (req->rl_connect_cookie == xprt->connect_cookie)
+ if (rqst->rq_connect_cookie == xprt->connect_cookie)
goto drop_connection;
- req->rl_connect_cookie = xprt->connect_cookie;
+ rqst->rq_xtime = ktime_get();
__set_bit(RPCRDMA_REQ_F_PENDING, &req->rl_flags);
if (rpcrdma_ep_post(&r_xprt->rx_ia, &r_xprt->rx_ep, req))
rqst->rq_xmit_bytes_sent += rqst->rq_snd_buf.len;
rqst->rq_bytes_sent = 0;
+
+ /* An RPC with no reply will throw off credit accounting,
+ * so drop the connection to reset the credit grant.
+ */
+ if (!rpc_reply_expected(task))
+ goto drop_connection;
return 0;
failed_marshal:
wait_for_completion(&ia->ri_remove_done);
ia->ri_id = NULL;
- ia->ri_pd = NULL;
ia->ri_device = NULL;
/* Return 1 to ensure the core destroys the id. */
return 1;
case RDMA_CM_EVENT_ESTABLISHED:
+ ++xprt->rx_xprt.connect_cookie;
connstate = 1;
rpcrdma_update_connect_private(xprt, &event->param.conn);
goto connected;
connstate = -EAGAIN;
goto connected;
case RDMA_CM_EVENT_DISCONNECTED:
+ ++xprt->rx_xprt.connect_cookie;
connstate = -ECONNABORTED;
connected:
xprt->rx_buf.rb_credits = 1;
ia->ri_id->qp = NULL;
}
ib_free_cq(ep->rep_attr.recv_cq);
+ ep->rep_attr.recv_cq = NULL;
ib_free_cq(ep->rep_attr.send_cq);
+ ep->rep_attr.send_cq = NULL;
/* The ULP is responsible for ensuring all DMA
* mappings and MRs are gone.
rpcrdma_dma_unmap_regbuf(req->rl_recvbuf);
}
rpcrdma_mrs_destroy(buf);
+ ib_dealloc_pd(ia->ri_pd);
+ ia->ri_pd = NULL;
/* Allow waiters to continue */
complete(&ia->ri_remove_done);
/* Client offers RDMA Read but does not initiate */
ep->rep_remote_cma.initiator_depth = 0;
- if (ia->ri_device->attrs.max_qp_rd_atom > 32) /* arbitrary but <= 255 */
- ep->rep_remote_cma.responder_resources = 32;
- else
- ep->rep_remote_cma.responder_resources =
- ia->ri_device->attrs.max_qp_rd_atom;
+ ep->rep_remote_cma.responder_resources =
+ min_t(int, U8_MAX, ia->ri_device->attrs.max_qp_rd_atom);
/* Limit transport retries so client can detect server
* GID changes quickly. RPC layer handles re-establishing
{
cancel_delayed_work_sync(&ep->rep_connect_worker);
- if (ia->ri_id->qp) {
+ if (ia->ri_id && ia->ri_id->qp) {
rpcrdma_ep_disconnect(ep, ia);
rdma_destroy_qp(ia->ri_id);
ia->ri_id->qp = NULL;
}
- ib_free_cq(ep->rep_attr.recv_cq);
- ib_free_cq(ep->rep_attr.send_cq);
+ if (ep->rep_attr.recv_cq)
+ ib_free_cq(ep->rep_attr.recv_cq);
+ if (ep->rep_attr.send_cq)
+ ib_free_cq(ep->rep_attr.send_cq);
}
/* Re-establish a connection after a device removal event.
LIST_HEAD(free);
LIST_HEAD(all);
- for (count = 0; count < 32; count++) {
+ for (count = 0; count < 3; count++) {
struct rpcrdma_mr *mr;
int rc;
list_splice(&all, &buf->rb_all);
r_xprt->rx_stats.mrs_allocated += count;
spin_unlock(&buf->rb_mrlock);
-
trace_xprtrdma_createmrs(r_xprt, count);
+
+ xprt_write_space(&r_xprt->rx_xprt);
}
static void
rpcrdma_create_req(struct rpcrdma_xprt *r_xprt)
{
struct rpcrdma_buffer *buffer = &r_xprt->rx_buf;
+ struct rpcrdma_regbuf *rb;
struct rpcrdma_req *req;
req = kzalloc(sizeof(*req), GFP_KERNEL);
if (req == NULL)
return ERR_PTR(-ENOMEM);
+ rb = rpcrdma_alloc_regbuf(RPCRDMA_HDRBUF_SIZE,
+ DMA_TO_DEVICE, GFP_KERNEL);
+ if (IS_ERR(rb)) {
+ kfree(req);
+ return ERR_PTR(-ENOMEM);
+ }
+ req->rl_rdmabuf = rb;
+ xdr_buf_init(&req->rl_hdrbuf, rb->rg_base, rdmab_length(rb));
+ req->rl_buffer = buffer;
+ INIT_LIST_HEAD(&req->rl_registered);
+
spin_lock(&buffer->rb_reqslock);
list_add(&req->rl_all, &buffer->rb_allreqs);
spin_unlock(&buffer->rb_reqslock);
- req->rl_buffer = &r_xprt->rx_buf;
- INIT_LIST_HEAD(&req->rl_registered);
return req;
}
struct rpcrdma_req *req)
{
struct ib_send_wr *send_wr = &req->rl_sendctx->sc_wr;
- struct ib_send_wr *send_wr_fail;
int rc;
if (req->rl_reply) {
--ep->rep_send_count;
}
- rc = ib_post_send(ia->ri_id->qp, send_wr, &send_wr_fail);
+ rc = ia->ri_ops->ro_send(ia, req);
trace_xprtrdma_post_send(req, rc);
if (rc)
return -ENOTCONN;
struct rpcrdma_buffer;
struct rpcrdma_req {
struct list_head rl_list;
- int rl_cpu;
- unsigned int rl_connect_cookie;
struct rpcrdma_buffer *rl_buffer;
struct rpcrdma_rep *rl_reply;
struct xdr_stream rl_stream;
(*ro_map)(struct rpcrdma_xprt *,
struct rpcrdma_mr_seg *, int, bool,
struct rpcrdma_mr **);
+ int (*ro_send)(struct rpcrdma_ia *ia,
+ struct rpcrdma_req *req);
void (*ro_reminv)(struct rpcrdma_rep *rep,
struct list_head *mrs);
void (*ro_unmap_sync)(struct rpcrdma_xprt *,
xs_pktdump("packet data:",
req->rq_svec->iov_base, req->rq_svec->iov_len);
+ req->rq_xtime = ktime_get();
status = xs_sendpages(transport->sock, NULL, 0, xdr, req->rq_bytes_sent,
true, &sent);
dprintk("RPC: %s(%u) = %d\n",
if (!xprt_bound(xprt))
return -ENOTCONN;
+ req->rq_xtime = ktime_get();
status = xs_sendpages(transport->sock, xs_addr(xprt), xprt->addrlen,
xdr, req->rq_bytes_sent, true, &sent);
/* Continue transmitting the packet/record. We must be careful
* to cope with writespace callbacks arriving _after_ we have
* called sendmsg(). */
+ req->rq_xtime = ktime_get();
while (1) {
sent = 0;
status = xs_sendpages(transport->sock, NULL, 0, xdr,
if (!rovr)
goto out_unlock;
xprt_pin_rqst(rovr);
+ xprt_update_rtt(rovr->rq_task);
spin_unlock(&xprt->recv_lock);
task = rovr->rq_task;