ceph: use i_ceph_lock instead of i_lock
We have been using i_lock to protect all kinds of data structures in the
ceph_inode_info struct, including lists of inodes that we need to iterate
over while avoiding races with inode destruction. That requires grabbing
a reference to the inode with the list lock protected, but igrab() now
takes i_lock to check the inode flags.
Changing the list lock ordering would be a painful process.
However, using a ceph-specific i_ceph_lock in the ceph inode instead of
i_lock is a simple mechanical change and avoids the ordering constraints
imposed by igrab().
Reported-by: Amon Ott <a.ott@m-privacy.de>
Signed-off-by: Sage Weil <sage@newdream.net>
diff --git a/fs/ceph/xattr.c b/fs/ceph/xattr.c
index 96c6739..a5e36e4 100644
--- a/fs/ceph/xattr.c
+++ b/fs/ceph/xattr.c
@@ -343,8 +343,8 @@
}
static int __build_xattrs(struct inode *inode)
- __releases(inode->i_lock)
- __acquires(inode->i_lock)
+ __releases(ci->i_ceph_lock)
+ __acquires(ci->i_ceph_lock)
{
u32 namelen;
u32 numattr = 0;
@@ -372,7 +372,7 @@
end = p + ci->i_xattrs.blob->vec.iov_len;
ceph_decode_32_safe(&p, end, numattr, bad);
xattr_version = ci->i_xattrs.version;
- spin_unlock(&inode->i_lock);
+ spin_unlock(&ci->i_ceph_lock);
xattrs = kcalloc(numattr, sizeof(struct ceph_xattr *),
GFP_NOFS);
@@ -387,7 +387,7 @@
goto bad_lock;
}
- spin_lock(&inode->i_lock);
+ spin_lock(&ci->i_ceph_lock);
if (ci->i_xattrs.version != xattr_version) {
/* lost a race, retry */
for (i = 0; i < numattr; i++)
@@ -418,7 +418,7 @@
return err;
bad_lock:
- spin_lock(&inode->i_lock);
+ spin_lock(&ci->i_ceph_lock);
bad:
if (xattrs) {
for (i = 0; i < numattr; i++)
@@ -512,7 +512,7 @@
if (vxattrs)
vxattr = ceph_match_vxattr(vxattrs, name);
- spin_lock(&inode->i_lock);
+ spin_lock(&ci->i_ceph_lock);
dout("getxattr %p ver=%lld index_ver=%lld\n", inode,
ci->i_xattrs.version, ci->i_xattrs.index_version);
@@ -520,14 +520,14 @@
(ci->i_xattrs.index_version >= ci->i_xattrs.version)) {
goto get_xattr;
} else {
- spin_unlock(&inode->i_lock);
+ spin_unlock(&ci->i_ceph_lock);
/* get xattrs from mds (if we don't already have them) */
err = ceph_do_getattr(inode, CEPH_STAT_CAP_XATTR);
if (err)
return err;
}
- spin_lock(&inode->i_lock);
+ spin_lock(&ci->i_ceph_lock);
if (vxattr && vxattr->readonly) {
err = vxattr->getxattr_cb(ci, value, size);
@@ -558,7 +558,7 @@
memcpy(value, xattr->val, xattr->val_len);
out:
- spin_unlock(&inode->i_lock);
+ spin_unlock(&ci->i_ceph_lock);
return err;
}
@@ -573,7 +573,7 @@
u32 len;
int i;
- spin_lock(&inode->i_lock);
+ spin_lock(&ci->i_ceph_lock);
dout("listxattr %p ver=%lld index_ver=%lld\n", inode,
ci->i_xattrs.version, ci->i_xattrs.index_version);
@@ -581,13 +581,13 @@
(ci->i_xattrs.index_version >= ci->i_xattrs.version)) {
goto list_xattr;
} else {
- spin_unlock(&inode->i_lock);
+ spin_unlock(&ci->i_ceph_lock);
err = ceph_do_getattr(inode, CEPH_STAT_CAP_XATTR);
if (err)
return err;
}
- spin_lock(&inode->i_lock);
+ spin_lock(&ci->i_ceph_lock);
err = __build_xattrs(inode);
if (err < 0)
@@ -619,7 +619,7 @@
}
out:
- spin_unlock(&inode->i_lock);
+ spin_unlock(&ci->i_ceph_lock);
return err;
}
@@ -739,7 +739,7 @@
if (!xattr)
goto out;
- spin_lock(&inode->i_lock);
+ spin_lock(&ci->i_ceph_lock);
retry:
issued = __ceph_caps_issued(ci, NULL);
if (!(issued & CEPH_CAP_XATTR_EXCL))
@@ -752,12 +752,12 @@
required_blob_size > ci->i_xattrs.prealloc_blob->alloc_len) {
struct ceph_buffer *blob = NULL;
- spin_unlock(&inode->i_lock);
+ spin_unlock(&ci->i_ceph_lock);
dout(" preaallocating new blob size=%d\n", required_blob_size);
blob = ceph_buffer_new(required_blob_size, GFP_NOFS);
if (!blob)
goto out;
- spin_lock(&inode->i_lock);
+ spin_lock(&ci->i_ceph_lock);
if (ci->i_xattrs.prealloc_blob)
ceph_buffer_put(ci->i_xattrs.prealloc_blob);
ci->i_xattrs.prealloc_blob = blob;
@@ -770,13 +770,13 @@
dirty = __ceph_mark_dirty_caps(ci, CEPH_CAP_XATTR_EXCL);
ci->i_xattrs.dirty = true;
inode->i_ctime = CURRENT_TIME;
- spin_unlock(&inode->i_lock);
+ spin_unlock(&ci->i_ceph_lock);
if (dirty)
__mark_inode_dirty(inode, dirty);
return err;
do_sync:
- spin_unlock(&inode->i_lock);
+ spin_unlock(&ci->i_ceph_lock);
err = ceph_sync_setxattr(dentry, name, value, size, flags);
out:
kfree(newname);
@@ -833,7 +833,7 @@
return -EOPNOTSUPP;
}
- spin_lock(&inode->i_lock);
+ spin_lock(&ci->i_ceph_lock);
__build_xattrs(inode);
issued = __ceph_caps_issued(ci, NULL);
dout("removexattr %p issued %s\n", inode, ceph_cap_string(issued));
@@ -846,12 +846,12 @@
ci->i_xattrs.dirty = true;
inode->i_ctime = CURRENT_TIME;
- spin_unlock(&inode->i_lock);
+ spin_unlock(&ci->i_ceph_lock);
if (dirty)
__mark_inode_dirty(inode, dirty);
return err;
do_sync:
- spin_unlock(&inode->i_lock);
+ spin_unlock(&ci->i_ceph_lock);
err = ceph_send_removexattr(dentry, name);
return err;
}