xfs: split metadata and log buffer completion to separate workqueues

XFS traditionally sends all buffer I/O completion work to a single
workqueue. This includes metadata buffer completion and log buffer
completion. The log buffer completion requires a high priority queue to
prevent stalls due to log forces getting stuck behind other queued work.

Rather than continue to prioritize all buffer I/O completion due to the
needs of log completion, split log buffer completion off to
m_log_workqueue and move the high priority flag from m_buf_workqueue to
m_log_workqueue.

Add a b_ioend_wq wq pointer to xfs_buf to allow completion workqueue
customization on a per-buffer basis. Initialize b_ioend_wq to
m_buf_workqueue by default in the generic buffer I/O submission path.
Finally, override the default wq with the high priority m_log_workqueue
in the log buffer I/O submission path.

Signed-off-by: Brian Foster <bfoster@redhat.com>
Reviewed-by: Dave Chinner <dchinner@redhat.com>
Signed-off-by: Dave Chinner <david@fromorbit.com>

4 files changed, 18 insertions, 7 deletions

diff --git a/fs/xfs/xfs_buf.c b/fs/xfs/xfs_buf.c
index d083889..945bea9 100644
--- a/fs/xfs/xfs_buf.c
+++ b/fs/xfs/xfs_buf.c
@@ -1041,7 +1041,7 @@ xfs_buf_ioend_work(
 	struct work_struct	*work)
 {
 	struct xfs_buf		*bp =
-		container_of(work, xfs_buf_t, b_iodone_work);
+		container_of(work, xfs_buf_t, b_ioend_work);
 
 	xfs_buf_ioend(bp);
 }
@@ -1050,8 +1050,8 @@ void
 xfs_buf_ioend_async(
 	struct xfs_buf	*bp)
 {
-	INIT_WORK(&bp->b_iodone_work, xfs_buf_ioend_work);
-	queue_work(bp->b_target->bt_mount->m_buf_workqueue, &bp->b_iodone_work);
+	INIT_WORK(&bp->b_ioend_work, xfs_buf_ioend_work);
+	queue_work(bp->b_ioend_wq, &bp->b_ioend_work);
 }
 
 void
@@ -1220,6 +1220,13 @@ _xfs_buf_ioapply(
 	 */
 	bp->b_error = 0;
 
+	/*
+	 * Initialize the I/O completion workqueue if we haven't yet or the
+	 * submitter has not opted to specify a custom one.
+	 */
+	if (!bp->b_ioend_wq)
+		bp->b_ioend_wq = bp->b_target->bt_mount->m_buf_workqueue;
+
 	if (bp->b_flags & XBF_WRITE) {
 		if (bp->b_flags & XBF_SYNCIO)
 			rw = WRITE_SYNC;
diff --git a/fs/xfs/xfs_buf.h b/fs/xfs/xfs_buf.h
index 82002c0..75ff5d5 100644
--- a/fs/xfs/xfs_buf.h
+++ b/fs/xfs/xfs_buf.h
@@ -164,7 +164,8 @@ typedef struct xfs_buf {
 	struct xfs_perag	*b_pag;		/* contains rbtree root */
 	xfs_buftarg_t		*b_target;	/* buffer target (device) */
 	void			*b_addr;	/* virtual address of buffer */
-	struct work_struct	b_iodone_work;
+	struct work_struct	b_ioend_work;
+	struct workqueue_struct	*b_ioend_wq;	/* I/O completion wq */
 	xfs_buf_iodone_t	b_iodone;	/* I/O completion function */
 	struct completion	b_iowait;	/* queue for I/O waiters */
 	void			*b_fspriv;
diff --git a/fs/xfs/xfs_log.c b/fs/xfs/xfs_log.c
index e810e9d..d508a65 100644
--- a/fs/xfs/xfs_log.c
+++ b/fs/xfs/xfs_log.c
@@ -1808,6 +1808,8 @@ xlog_sync(
 	XFS_BUF_ZEROFLAGS(bp);
 	XFS_BUF_ASYNC(bp);
 	bp->b_flags |= XBF_SYNCIO;
+	/* use high priority completion wq */
+	bp->b_ioend_wq = log->l_mp->m_log_workqueue;
 
 	if (log->l_mp->m_flags & XFS_MOUNT_BARRIER) {
 		bp->b_flags |= XBF_FUA;
@@ -1856,6 +1858,8 @@ xlog_sync(
 		bp->b_flags |= XBF_SYNCIO;
 		if (log->l_mp->m_flags & XFS_MOUNT_BARRIER)
 			bp->b_flags |= XBF_FUA;
+		/* use high priority completion wq */
+		bp->b_ioend_wq = log->l_mp->m_log_workqueue;
 
 		ASSERT(XFS_BUF_ADDR(bp) <= log->l_logBBsize-1);
 		ASSERT(XFS_BUF_ADDR(bp) + BTOBB(count) <= log->l_logBBsize);
diff --git a/fs/xfs/xfs_super.c b/fs/xfs/xfs_super.c
index 45fa5b5..b5aacfc 100644
--- a/fs/xfs/xfs_super.c
+++ b/fs/xfs/xfs_super.c
@@ -842,8 +842,7 @@ xfs_init_mount_workqueues(
 	struct xfs_mount	*mp)
 {
 	mp->m_buf_workqueue = alloc_workqueue("xfs-buf/%s",
-			WQ_MEM_RECLAIM|WQ_HIGHPRI|WQ_FREEZABLE, 1,
-			mp->m_fsname);
+			WQ_MEM_RECLAIM|WQ_FREEZABLE, 1, mp->m_fsname);
 	if (!mp->m_buf_workqueue)
 		goto out;
 
@@ -868,7 +867,7 @@ xfs_init_mount_workqueues(
 		goto out_destroy_cil;
 
 	mp->m_log_workqueue = alloc_workqueue("xfs-log/%s",
-			WQ_FREEZABLE, 0, mp->m_fsname);
+			WQ_FREEZABLE|WQ_HIGHPRI, 0, mp->m_fsname);
 	if (!mp->m_log_workqueue)
 		goto out_destroy_reclaim;