xfs: exclude never-released buffers from buftarg I/O accounting

The upcoming buftarg I/O accounting mechanism maintains a count of
all buffers that have undergone I/O in the current hold-release
cycle.  Certain buffers associated with core infrastructure (e.g.,
the xfs_mount superblock buffer, log buffers) are never released,
however. This means that accounting I/O submission on such buffers
elevates the buftarg count indefinitely and could lead to lockup on
unmount.

Define a new buffer flag to explicitly exclude buffers from buftarg
I/O accounting. Set the flag on the superblock and associated log
buffers.

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

diff --git a/fs/xfs/xfs_buf.c b/fs/xfs/xfs_buf.c
index 5d52e44..c0bd5e0 100644 (file)
--- a/fs/xfs/xfs_buf.c
+++ b/fs/xfs/xfs_buf.c
@@ -815,7 +815,8 @@ xfs_buf_get_uncached(
        struct xfs_buf          *bp;
        DEFINE_SINGLE_BUF_MAP(map, XFS_BUF_DADDR_NULL, numblks);
 
-       bp = _xfs_buf_alloc(target, &map, 1, 0);
+       /* flags might contain irrelevant bits, pass only what we care about */
+       bp = _xfs_buf_alloc(target, &map, 1, flags & XBF_NO_IOACCT);
        if (unlikely(bp == NULL))
                goto fail;
 

diff --git a/fs/xfs/xfs_buf.h b/fs/xfs/xfs_buf.h
index 8bfb974..e2108da 100644 (file)
--- a/fs/xfs/xfs_buf.h
+++ b/fs/xfs/xfs_buf.h
@@ -43,6 +43,7 @@ typedef enum {
 #define XBF_READ        (1 << 0) /* buffer intended for reading from device */
 #define XBF_WRITE       (1 << 1) /* buffer intended for writing to device */
 #define XBF_READ_AHEAD  (1 << 2) /* asynchronous read-ahead */
+#define XBF_NO_IOACCT   (1 << 3) /* bypass I/O accounting (non-LRU bufs) */
 #define XBF_ASYNC       (1 << 4) /* initiator will not wait for completion */
 #define XBF_DONE        (1 << 5) /* all pages in the buffer uptodate */
 #define XBF_STALE       (1 << 6) /* buffer has been staled, do not find it */

diff --git a/fs/xfs/xfs_log.c b/fs/xfs/xfs_log.c
index bde02f1..216aaa2 100644 (file)
--- a/fs/xfs/xfs_log.c
+++ b/fs/xfs/xfs_log.c
@@ -1415,7 +1415,7 @@ xlog_alloc_log(
         */
        error = -ENOMEM;
        bp = xfs_buf_alloc(mp->m_logdev_targp, XFS_BUF_DADDR_NULL,
-                          BTOBB(log->l_iclog_size), 0);
+                          BTOBB(log->l_iclog_size), XBF_NO_IOACCT);
        if (!bp)
                goto out_free_log;
 
@@ -1454,7 +1454,8 @@ xlog_alloc_log(
                prev_iclog = iclog;
 
                bp = xfs_buf_get_uncached(mp->m_logdev_targp,
-                                               BTOBB(log->l_iclog_size), 0);
+                                         BTOBB(log->l_iclog_size),
+                                         XBF_NO_IOACCT);
                if (!bp)
                        goto out_free_iclog;
 

diff --git a/fs/xfs/xfs_mount.c b/fs/xfs/xfs_mount.c
index e39b023..970c19b 100644 (file)
--- a/fs/xfs/xfs_mount.c
+++ b/fs/xfs/xfs_mount.c
@@ -272,13 +272,15 @@ xfs_readsb(
        buf_ops = NULL;
 
        /*
-        * Allocate a (locked) buffer to hold the superblock.
-        * This will be kept around at all times to optimize
-        * access to the superblock.
+        * Allocate a (locked) buffer to hold the superblock. This will be kept
+        * around at all times to optimize access to the superblock. Therefore,
+        * set XBF_NO_IOACCT to make sure it doesn't hold the buftarg count
+        * elevated.
         */
 reread:
        error = xfs_buf_read_uncached(mp->m_ddev_targp, XFS_SB_DADDR,
-                                  BTOBB(sector_size), 0, &bp, buf_ops);
+                                     BTOBB(sector_size), XBF_NO_IOACCT, &bp,
+                                     buf_ops);
        if (error) {
                if (loud)
                        xfs_warn(mp, "SB validate failed with error %d.", error);