jbd2: Fix I/O hang in jbd2_journal_release_jbd_inode

This fixes a hang seen in jbd2_journal_release_jbd_inode
on a lot of Power 6 systems running with ext4. When we get
in the hung state, all I/O to the disk in question gets blocked
where we stay indefinitely. Looking at the task list, I can see
we are stuck in jbd2_journal_release_jbd_inode waiting on a
wake up. I added some debug code to detect this scenario and
dump additional data if we were stuck in jbd2_journal_release_jbd_inode
for longer than 30 minutes. When it hit, I was able to see that
i_flags was 0, suggesting we missed the wake up.

This patch changes i_flags to be an unsigned long, uses bit operators
to access it, and adds barriers around the accesses. Prior to applying
this patch, we were regularly hitting this hang on numerous systems
in our test environment. After applying the patch, the hangs no longer
occur.

Signed-off-by: Brian King <brking@linux.vnet.ibm.com>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>

2 files changed, 11 insertions, 5 deletions

diff --git a/fs/jbd2/commit.c b/fs/jbd2/commit.c
index 7c068c1..6494c81 100644
--- a/fs/jbd2/commit.c
+++ b/fs/jbd2/commit.c
@@ -26,7 +26,9 @@
 #include <linux/backing-dev.h>
 #include <linux/bio.h>
 #include <linux/blkdev.h>
+#include <linux/bitops.h>
 #include <trace/events/jbd2.h>
+#include <asm/system.h>
 
 /*
  * Default IO end handler for temporary BJ_IO buffer_heads.
@@ -236,7 +238,7 @@ static int journal_submit_data_buffers(journal_t *journal,
 	spin_lock(&journal->j_list_lock);
 	list_for_each_entry(jinode, &commit_transaction->t_inode_list, i_list) {
 		mapping = jinode->i_vfs_inode->i_mapping;
-		jinode->i_flags |= JI_COMMIT_RUNNING;
+		set_bit(__JI_COMMIT_RUNNING, &jinode->i_flags);
 		spin_unlock(&journal->j_list_lock);
 		/*
 		 * submit the inode data buffers. We use writepage
@@ -251,7 +253,8 @@ static int journal_submit_data_buffers(journal_t *journal,
 		spin_lock(&journal->j_list_lock);
 		J_ASSERT(jinode->i_transaction == commit_transaction);
 		commit_transaction->t_flushed_data_blocks = 1;
-		jinode->i_flags &= ~JI_COMMIT_RUNNING;
+		clear_bit(__JI_COMMIT_RUNNING, &jinode->i_flags);
+		smp_mb__after_clear_bit();
 		wake_up_bit(&jinode->i_flags, __JI_COMMIT_RUNNING);
 	}
 	spin_unlock(&journal->j_list_lock);
@@ -272,7 +275,7 @@ static int journal_finish_inode_data_buffers(journal_t *journal,
 	/* For locking, see the comment in journal_submit_data_buffers() */
 	spin_lock(&journal->j_list_lock);
 	list_for_each_entry(jinode, &commit_transaction->t_inode_list, i_list) {
-		jinode->i_flags |= JI_COMMIT_RUNNING;
+		set_bit(__JI_COMMIT_RUNNING, &jinode->i_flags);
 		spin_unlock(&journal->j_list_lock);
 		err = filemap_fdatawait(jinode->i_vfs_inode->i_mapping);
 		if (err) {
@@ -288,7 +291,8 @@ static int journal_finish_inode_data_buffers(journal_t *journal,
 				ret = err;
 		}
 		spin_lock(&journal->j_list_lock);
-		jinode->i_flags &= ~JI_COMMIT_RUNNING;
+		clear_bit(__JI_COMMIT_RUNNING, &jinode->i_flags);
+		smp_mb__after_clear_bit();
 		wake_up_bit(&jinode->i_flags, __JI_COMMIT_RUNNING);
 	}
 
diff --git a/fs/jbd2/journal.c b/fs/jbd2/journal.c
index 0e8014e..75e1b5a 100644
--- a/fs/jbd2/journal.c
+++ b/fs/jbd2/journal.c
@@ -42,12 +42,14 @@
 #include <linux/log2.h>
 #include <linux/vmalloc.h>
 #include <linux/backing-dev.h>
+#include <linux/bitops.h>
 
 #define CREATE_TRACE_POINTS
 #include <trace/events/jbd2.h>
 
 #include <asm/uaccess.h>
 #include <asm/page.h>
+#include <asm/system.h>
 
 EXPORT_SYMBOL(jbd2_journal_extend);
 EXPORT_SYMBOL(jbd2_journal_stop);
@@ -2206,7 +2208,7 @@ void jbd2_journal_release_jbd_inode(journal_t *journal,
 restart:
 	spin_lock(&journal->j_list_lock);
 	/* Is commit writing out inode - we have to wait */
-	if (jinode->i_flags & JI_COMMIT_RUNNING) {
+	if (test_bit(__JI_COMMIT_RUNNING, &jinode->i_flags)) {
 		wait_queue_head_t *wq;
 		DEFINE_WAIT_BIT(wait, &jinode->i_flags, __JI_COMMIT_RUNNING);
 		wq = bit_waitqueue(&jinode->i_flags, __JI_COMMIT_RUNNING);