GFS2: Alter point of entry to glock lru list for glocks with an address_space
Rather than allowing the glocks to be scheduled for possible
reclaim as soon as they have exited the journal, this patch
delays their entry to the list until the glocks in question
are no longer in use.
This means that we will rely on the vm for writeback of all
dirty data and metadata from now on. When glocks are added
to the lru list they should be freeable much faster since all
the I/O required to free them should have already been completed.
This should lead to much better I/O patterns under low memory
conditions.
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
diff --git a/fs/gfs2/lops.c b/fs/gfs2/lops.c
index 51d27f0..611a51d 100644
--- a/fs/gfs2/lops.c
+++ b/fs/gfs2/lops.c
@@ -40,7 +40,7 @@
{
struct gfs2_bufdata *bd;
- gfs2_assert_withdraw(sdp, test_bit(SDF_JOURNAL_LIVE, &sdp->sd_flags));
+ BUG_ON(!current->journal_info);
clear_buffer_dirty(bh);
if (test_set_buffer_pinned(bh))
@@ -65,6 +65,7 @@
* @sdp: the filesystem the buffer belongs to
* @bh: The buffer to unpin
* @ai:
+ * @flags: The inode dirty flags
*
*/
@@ -73,10 +74,8 @@
{
struct gfs2_bufdata *bd = bh->b_private;
- gfs2_assert_withdraw(sdp, buffer_uptodate(bh));
-
- if (!buffer_pinned(bh))
- gfs2_assert_withdraw(sdp, 0);
+ BUG_ON(!buffer_uptodate(bh));
+ BUG_ON(!buffer_pinned(bh));
lock_buffer(bh);
mark_buffer_dirty(bh);
@@ -95,8 +94,7 @@
list_add(&bd->bd_ail_st_list, &ai->ai_ail1_list);
spin_unlock(&sdp->sd_ail_lock);
- if (test_and_clear_bit(GLF_LFLUSH, &bd->bd_gl->gl_flags))
- gfs2_glock_schedule_for_reclaim(bd->bd_gl);
+ clear_bit(GLF_LFLUSH, &bd->bd_gl->gl_flags);
trace_gfs2_pin(bd, 0);
unlock_buffer(bh);
atomic_dec(&sdp->sd_log_pinned);