From: Matthew Treinish <treinish@linux.vnet.ibm.com>
To: linux-nfs@vger.kernel.org
Cc: treinish@linux.vnet.ibm.com
Subject: [PATCH/RFC v3 09/10] VFH recovery from a replication/migration event.
Date: Mon, 5 Mar 2012 15:26:50 -0500 [thread overview]
Message-ID: <1330979211-894-10-git-send-email-treinish@linux.vnet.ibm.com> (raw)
In-Reply-To: <1330979211-894-1-git-send-email-treinish@linux.vnet.ibm.com>
Implements a refreshing file handles after a successful migration
or replication event. After the client changes to using a different
server if FH4_VOL_MIGRATION is set by the server it is assumed
that the file handles have all expired during migration or
replication.
nfs4_vfh_replication uses the dcache to find each file handle that
is currently cached and refreshes it using an nfs4 lookup.
Signed-off-by: Matthew Treinish <treinish@linux.vnet.ibm.com>
---
fs/nfs/nfs4_fs.h | 1 +
fs/nfs/nfs4proc.c | 73 +++++++++++++++++++++++++++++++++++++++++++++++++++++
2 files changed, 74 insertions(+), 0 deletions(-)
diff --git a/fs/nfs/nfs4_fs.h b/fs/nfs/nfs4_fs.h
index 52d7f94..b4ca67f 100644
--- a/fs/nfs/nfs4_fs.h
+++ b/fs/nfs/nfs4_fs.h
@@ -218,6 +218,7 @@ extern int nfs4_do_close(struct nfs4_state *state, gfp_t gfp_mask, int wait, boo
extern int nfs4_server_capabilities(struct nfs_server *server, struct nfs_fh *fhandle);
extern int nfs4_proc_fs_locations(struct inode *dir, const struct qstr *name,
struct nfs4_fs_locations *fs_locations, struct page *page);
+extern int nfs4_vfh_replication(struct dentry *d_root);
extern void nfs4_release_lockowner(const struct nfs4_lock_state *);
extern const struct xattr_handler *nfs4_xattr_handlers[];
diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c
index 78ed5c5..f9484c3 100644
--- a/fs/nfs/nfs4proc.c
+++ b/fs/nfs/nfs4proc.c
@@ -2558,6 +2558,79 @@ static int nfs4_proc_lookup(struct rpc_clnt *clnt, struct inode *dir, struct qst
return err;
}
+int nfs4_vfh_replication(struct dentry *d_root)
+{
+ struct nfs_fsinfo info;
+ int err = 0;
+ struct nfs_fattr *fattr = NULL;
+ struct dentry *this_parent;
+ struct list_head *next;
+ struct nfs_server *server = NFS_SB(d_root->d_sb);
+ fattr = nfs_alloc_fattr();
+ if (fattr == NULL)
+ return -ENOMEM;
+ info.fattr = fattr;
+ /* Check if we need to refresh filehandles */
+ if ((server->fhexpiretype & NFS4_FH_VOL_MIGRATION
+ || server->fhexpiretype & NFS4_FH_VOLATILE_ANY)
+ && server->flags & NFS_MOUNT_VFHRETRY) {
+ /* Update root dentry */
+ err = nfs4_proc_get_root(server, NFS_FH(d_root->d_inode),
+ &info);
+ if (!err) {
+ if (NFS_FILEID(d_root->d_inode) != info.fattr->fileid)
+ set_nfs_fileid(d_root->d_inode,
+ info.fattr->fileid);
+ nfs_copy_fh(server->rootfh, NFS_FH(d_root->d_inode));
+ /* Only needed if fsid changes on server */
+ memcpy(&server->fsid, &info.fattr->fsid,
+ sizeof(server->fsid));
+ } else
+ return err;
+ /* Loop over dcache entries */
+ this_parent = d_root;
+repeat:
+ next = this_parent->d_subdirs.next;
+resume:
+ while (next != &this_parent->d_subdirs) {
+ struct list_head *tmp = next;
+ struct dentry *dentry = list_entry(tmp, struct dentry,
+ d_u.d_child);
+ next = tmp->next;
+ /* Perform NFS lookup on dentry */
+ if (dentry->d_parent->d_inode && dentry->d_inode &&
+ &dentry->d_name)
+ nfs4_proc_lookup(server->client,
+ dentry->d_parent->d_inode,
+ &dentry->d_name,
+ NFS_FH(dentry->d_inode),
+ fattr);
+ if (!list_empty(&dentry->d_subdirs)) {
+ this_parent = dentry;
+ goto repeat;
+ }
+ }
+ if (this_parent != d_root) {
+ struct dentry *child = this_parent;
+ struct dentry *new = this_parent->d_parent;
+ /*
+ * might go back up the wrong parent if we have had a rename
+ * or deletion return ESTALE if it is renamed or deleted
+ */
+ if (new != this_parent->d_parent ||
+ (this_parent->d_flags & DCACHE_DISCONNECTED)) {
+ return -ESTALE;
+ }
+ this_parent = new;
+ next = child->d_u.d_child.next;
+ goto resume;
+ }
+ return 0;
+
+ } else
+ return 0;
+}
+
/* In the case of gettattr only a fh is given, and we need a VFS object to do
* vfh recovery. This function traverses the dcache to find the dentry with
* a matching fh
--
1.7.4.4
next prev parent reply other threads:[~2012-03-05 20:27 UTC|newest]
Thread overview: 12+ messages / expand[flat|nested] mbox.gz Atom feed top
2012-03-05 20:26 [PATCH/RFC v3 00/10] Volatile File Handle Client-side Support Matthew Treinish
2012-03-05 20:26 ` [PATCH/RFC v3 01/10] Add support for FH_EXPIRE_TYPE attribute Matthew Treinish
2012-03-05 20:26 ` [PATCH/RFC v3 02/10] Save root file handle in nfs_server Matthew Treinish
2012-03-05 20:26 ` [PATCH/RFC v3 03/10] Store root dentry in server object Matthew Treinish
2012-03-05 20:26 ` [PATCH/RFC v3 04/10] Store objects in nfs4_exception to be used during FHEXPIRED recovery Matthew Treinish
2012-03-05 20:26 ` [PATCH/RFC v3 05/10] Add nfs4_vfh_getdentry() for getattr recovery Matthew Treinish
2012-03-05 20:26 ` [PATCH/RFC v3 06/10] Add VFH FHEXPIRED recovery functions Matthew Treinish
2012-03-05 20:26 ` [PATCH/RFC v3 07/10] Perform recovery on both inodes for rename Matthew Treinish
2012-03-05 20:26 ` [PATCH/RFC v3 08/10] Added error handling for NFS4ERR_FHEXPIRED Matthew Treinish
2012-03-05 20:26 ` Matthew Treinish [this message]
2012-03-05 20:26 ` [PATCH/RFC v3 10/10] New mount option for volatile filehandle recovery Matthew Treinish
2012-03-13 17:33 ` [PATCH/RFC v3 00/10] Volatile File Handle Client-side Support Matthew Treinish
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=1330979211-894-10-git-send-email-treinish@linux.vnet.ibm.com \
--to=treinish@linux.vnet.ibm.com \
--cc=linux-nfs@vger.kernel.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).