ceph: record truncate size/seq for snap data writeback
Dirty snapshot data needs to be flushed unconditionally. If they
were created before truncation, writeback should use old truncate
size/seq.
Signed-off-by: Yan, Zheng <zyan@redhat.com>
diff --git a/fs/ceph/addr.c b/fs/ceph/addr.c
index dbb5f7d..7d69f25 100644
--- a/fs/ceph/addr.c
+++ b/fs/ceph/addr.c
@@ -474,7 +474,9 @@
* only snap context we are allowed to write back.
*/
static struct ceph_snap_context *get_oldest_context(struct inode *inode,
- loff_t *snap_size)
+ loff_t *snap_size,
+ u64 *truncate_size,
+ u32 *truncate_seq)
{
struct ceph_inode_info *ci = ceph_inode(inode);
struct ceph_snap_context *snapc = NULL;
@@ -488,6 +490,10 @@
snapc = ceph_get_snap_context(capsnap->context);
if (snap_size)
*snap_size = capsnap->size;
+ if (truncate_size)
+ *truncate_size = capsnap->truncate_size;
+ if (truncate_seq)
+ *truncate_seq = capsnap->truncate_seq;
break;
}
}
@@ -495,6 +501,10 @@
snapc = ceph_get_snap_context(ci->i_head_snapc);
dout(" head snapc %p has %d dirty pages\n",
snapc, ci->i_wrbuffer_ref_head);
+ if (truncate_size)
+ *truncate_size = capsnap->truncate_size;
+ if (truncate_seq)
+ *truncate_seq = capsnap->truncate_seq;
}
spin_unlock(&ci->i_ceph_lock);
return snapc;
@@ -537,7 +547,8 @@
dout("writepage %p page %p not dirty?\n", inode, page);
goto out;
}
- oldest = get_oldest_context(inode, &snap_size);
+ oldest = get_oldest_context(inode, &snap_size,
+ &truncate_size, &truncate_seq);
if (snapc->seq > oldest->seq) {
dout("writepage %p page %p snapc %p not writeable - noop\n",
inode, page, snapc);
@@ -548,12 +559,8 @@
}
ceph_put_snap_context(oldest);
- spin_lock(&ci->i_ceph_lock);
- truncate_seq = ci->i_truncate_seq;
- truncate_size = ci->i_truncate_size;
if (snap_size == -1)
snap_size = i_size_read(inode);
- spin_unlock(&ci->i_ceph_lock);
/* is this a partial page at end of file? */
if (page_off >= snap_size) {
@@ -800,7 +807,8 @@
/* find oldest snap context with dirty data */
ceph_put_snap_context(snapc);
snap_size = -1;
- snapc = get_oldest_context(inode, &snap_size);
+ snapc = get_oldest_context(inode, &snap_size,
+ &truncate_size, &truncate_seq);
if (!snapc) {
/* hmm, why does writepages get called when there
is no dirty data? */
@@ -810,11 +818,7 @@
dout(" oldest snapc is %p seq %lld (%d snaps)\n",
snapc, snapc->seq, snapc->num_snaps);
- spin_lock(&ci->i_ceph_lock);
- truncate_seq = ci->i_truncate_seq;
- truncate_size = ci->i_truncate_size;
i_size = i_size_read(inode);
- spin_unlock(&ci->i_ceph_lock);
if (last_snapc && snapc != last_snapc) {
/* if we switched to a newer snapc, restart our scan at the
@@ -1160,7 +1164,8 @@
static int context_is_writeable_or_written(struct inode *inode,
struct ceph_snap_context *snapc)
{
- struct ceph_snap_context *oldest = get_oldest_context(inode, NULL);
+ struct ceph_snap_context *oldest = get_oldest_context(inode, NULL,
+ NULL, NULL);
int ret = !oldest || snapc->seq <= oldest->seq;
ceph_put_snap_context(oldest);
@@ -1205,7 +1210,7 @@
* this page is already dirty in another (older) snap
* context! is it writeable now?
*/
- oldest = get_oldest_context(inode, NULL);
+ oldest = get_oldest_context(inode, NULL, NULL, NULL);
if (snapc->seq > oldest->seq) {
ceph_put_snap_context(oldest);
diff --git a/fs/ceph/snap.c b/fs/ceph/snap.c
index 9ff5219..8f8b41c 100644
--- a/fs/ceph/snap.c
+++ b/fs/ceph/snap.c
@@ -593,6 +593,8 @@
capsnap->atime = inode->i_atime;
capsnap->ctime = inode->i_ctime;
capsnap->time_warp_seq = ci->i_time_warp_seq;
+ capsnap->truncate_size = ci->i_truncate_size;
+ capsnap->truncate_seq = ci->i_truncate_seq;
if (capsnap->dirty_pages) {
dout("finish_cap_snap %p cap_snap %p snapc %p %llu %s s=%llu "
"still has %d dirty pages\n", inode, capsnap,
diff --git a/fs/ceph/super.h b/fs/ceph/super.h
index b07f55e..1b31ec6 100644
--- a/fs/ceph/super.h
+++ b/fs/ceph/super.h
@@ -181,6 +181,8 @@
u64 size;
struct timespec mtime, atime, ctime;
u64 time_warp_seq;
+ u64 truncate_size;
+ u32 truncate_seq;
int writing; /* a sync write is still in progress */
int dirty_pages; /* dirty pages awaiting writeback */
bool inline_data;