GFS2: update freeze code to use freeze/thaw_super on all nodes

The current gfs2 freezing code is considerably more complicated than it
should be because it doesn't use the vfs freezing code on any node except
the one that begins the freeze.  This is because it needs to acquire a
cluster glock before calling the vfs code to prevent a deadlock, and
without the new freeze_super and thaw_super hooks, that was impossible. To
deal with the issue, gfs2 had to do some hacky locking tricks to make sure
that a frozen node couldn't be holding on a lock it needed to do the
unfreeze ioctl.

This patch makes use of the new hooks to simply the gfs2 locking code. Now,
all the nodes in the cluster freeze and thaw in exactly the same way. Every
node in the cluster caches the freeze glock in the shared state.  The new
freeze_super hook allows the freezing node to grab this freeze glock in
the exclusive state without first calling the vfs freeze_super function.
All the nodes in the cluster see this lock change, and call the vfs
freeze_super function. The vfs locking code guarantees that the nodes can't
get stuck holding the glocks necessary to unfreeze the system.  To
unfreeze, the freezing node uses the new thaw_super hook to drop the freeze
glock. Again, all the nodes notice this, reacquire the glock in shared mode
and call the vfs thaw_super function.

Signed-off-by: Benjamin Marzinski <bmarzins@redhat.com>
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>

1 files changed, 21 insertions, 21 deletions

diff --git a/fs/gfs2/log.c b/fs/gfs2/log.c
index 3966fad..536e7a6 100644
--- a/fs/gfs2/log.c
+++ b/fs/gfs2/log.c
@@ -339,6 +339,7 @@ void gfs2_log_release(struct gfs2_sbd *sdp, unsigned int blks)
 
 int gfs2_log_reserve(struct gfs2_sbd *sdp, unsigned int blks)
 {
+	int ret = 0;
 	unsigned reserved_blks = 7 * (4096 / sdp->sd_vfs->s_blocksize);
 	unsigned wanted = blks + reserved_blks;
 	DEFINE_WAIT(wait);
@@ -362,9 +363,13 @@ retry:
 		} while(free_blocks <= wanted);
 		finish_wait(&sdp->sd_log_waitq, &wait);
 	}
+	atomic_inc(&sdp->sd_reserving_log);
 	if (atomic_cmpxchg(&sdp->sd_log_blks_free, free_blocks,
-				free_blocks - blks) != free_blocks)
+				free_blocks - blks) != free_blocks) {
+		if (atomic_dec_and_test(&sdp->sd_reserving_log))
+			wake_up(&sdp->sd_reserving_log_wait);
 		goto retry;
+	}
 	trace_gfs2_log_blocks(sdp, -blks);
 
 	/*
@@ -377,9 +382,11 @@ retry:
 	down_read(&sdp->sd_log_flush_lock);
 	if (unlikely(!test_bit(SDF_JOURNAL_LIVE, &sdp->sd_flags))) {
 		gfs2_log_release(sdp, blks);
-		return -EROFS;
+		ret = -EROFS;
 	}
-	return 0;
+	if (atomic_dec_and_test(&sdp->sd_reserving_log))
+		wake_up(&sdp->sd_reserving_log_wait);
+	return ret;
 }
 
 /**
@@ -652,9 +659,12 @@ static void log_write_header(struct gfs2_sbd *sdp, u32 flags)
 	u32 hash;
 	int rw = WRITE_FLUSH_FUA | REQ_META;
 	struct page *page = mempool_alloc(gfs2_page_pool, GFP_NOIO);
+	enum gfs2_freeze_state state = atomic_read(&sdp->sd_freeze_state);
 	lh = page_address(page);
 	clear_page(lh);
 
+	gfs2_assert_withdraw(sdp, (state != SFS_FROZEN));
+
 	tail = current_tail(sdp);
 
 	lh->lh_header.mh_magic = cpu_to_be32(GFS2_MAGIC);
@@ -695,6 +705,7 @@ void gfs2_log_flush(struct gfs2_sbd *sdp, struct gfs2_glock *gl,
 		    enum gfs2_flush_type type)
 {
 	struct gfs2_trans *tr;
+	enum gfs2_freeze_state state = atomic_read(&sdp->sd_freeze_state);
 
 	down_write(&sdp->sd_log_flush_lock);
 
@@ -713,8 +724,12 @@ void gfs2_log_flush(struct gfs2_sbd *sdp, struct gfs2_glock *gl,
 		INIT_LIST_HEAD(&tr->tr_ail1_list);
 		INIT_LIST_HEAD(&tr->tr_ail2_list);
 		tr->tr_first = sdp->sd_log_flush_head;
+		if (unlikely (state == SFS_FROZEN))
+			gfs2_assert_withdraw(sdp, !tr->tr_num_buf_new && !tr->tr_num_databuf_new);
 	}
 
+	if (unlikely(state == SFS_FROZEN))
+		gfs2_assert_withdraw(sdp, !sdp->sd_log_num_revoke);
 	gfs2_assert_withdraw(sdp,
 			sdp->sd_log_num_revoke == sdp->sd_log_commited_revoke);
 
@@ -745,8 +760,6 @@ void gfs2_log_flush(struct gfs2_sbd *sdp, struct gfs2_glock *gl,
 	spin_unlock(&sdp->sd_ail_lock);
 	gfs2_log_unlock(sdp);
 
-	if (atomic_read(&sdp->sd_log_freeze))
-		type = FREEZE_FLUSH;
 	if (type != NORMAL_FLUSH) {
 		if (!sdp->sd_log_idle) {
 			for (;;) {
@@ -763,21 +776,8 @@ void gfs2_log_flush(struct gfs2_sbd *sdp, struct gfs2_glock *gl,
 		}
 		if (type == SHUTDOWN_FLUSH || type == FREEZE_FLUSH)
 			gfs2_log_shutdown(sdp);
-		if (type == FREEZE_FLUSH) {
-			int error;
-
-			atomic_set(&sdp->sd_log_freeze, 0);
-			wake_up(&sdp->sd_log_frozen_wait);
-			error = gfs2_glock_nq_init(sdp->sd_freeze_gl,
-						   LM_ST_SHARED, 0,
-						   &sdp->sd_thaw_gh);
-			if (error) {
-				printk(KERN_INFO "GFS2: couln't get freeze lock : %d\n", error);
-				gfs2_assert_withdraw(sdp, 0);
-			}
-			else
-				gfs2_glock_dq_uninit(&sdp->sd_thaw_gh);
-		}
+		if (type == FREEZE_FLUSH)
+			atomic_set(&sdp->sd_freeze_state, SFS_FROZEN);
 	}
 
 	trace_gfs2_log_flush(sdp, 0);
@@ -888,7 +888,7 @@ void gfs2_log_shutdown(struct gfs2_sbd *sdp)
 
 static inline int gfs2_jrnl_flush_reqd(struct gfs2_sbd *sdp)
 {
-	return (atomic_read(&sdp->sd_log_pinned) >= atomic_read(&sdp->sd_log_thresh1) || atomic_read(&sdp->sd_log_freeze));
+	return (atomic_read(&sdp->sd_log_pinned) >= atomic_read(&sdp->sd_log_thresh1));
 }
 
 static inline int gfs2_ail_flush_reqd(struct gfs2_sbd *sdp)