xfs: Don't wake the aild once per second
[safe/jmp/linux-2.6] / fs / xfs / xfs_trans_ail.c
index 4d6330e..063dfbd 100644 (file)
@@ -1,5 +1,6 @@
 /*
  * Copyright (c) 2000-2002,2005 Silicon Graphics, Inc.
+ * Copyright (c) 2008 Dave Chinner
  * All Rights Reserved.
  *
  * This program is free software; you can redistribute it and/or
 #include "xfs_trans_priv.h"
 #include "xfs_error.h"
 
-STATIC void xfs_ail_insert(xfs_ail_entry_t *, xfs_log_item_t *);
-STATIC xfs_log_item_t * xfs_ail_delete(xfs_ail_entry_t *, xfs_log_item_t *);
-STATIC xfs_log_item_t * xfs_ail_min(xfs_ail_entry_t *);
-STATIC xfs_log_item_t * xfs_ail_next(xfs_ail_entry_t *, xfs_log_item_t *);
+STATIC void xfs_ail_insert(struct xfs_ail *, xfs_log_item_t *);
+STATIC xfs_log_item_t * xfs_ail_delete(struct xfs_ail *, xfs_log_item_t *);
+STATIC xfs_log_item_t * xfs_ail_min(struct xfs_ail *);
+STATIC xfs_log_item_t * xfs_ail_next(struct xfs_ail *, xfs_log_item_t *);
 
 #ifdef DEBUG
-STATIC void xfs_ail_check(xfs_ail_entry_t *, xfs_log_item_t *);
+STATIC void xfs_ail_check(struct xfs_ail *, xfs_log_item_t *);
 #else
 #define        xfs_ail_check(a,l)
 #endif /* DEBUG */
@@ -50,20 +51,20 @@ STATIC void xfs_ail_check(xfs_ail_entry_t *, xfs_log_item_t *);
  * lsn of the last item in the AIL.
  */
 xfs_lsn_t
-xfs_trans_tail_ail(
-       xfs_mount_t     *mp)
+xfs_trans_ail_tail(
+       struct xfs_ail  *ailp)
 {
        xfs_lsn_t       lsn;
        xfs_log_item_t  *lip;
 
-       spin_lock(&mp->m_ail_lock);
-       lip = xfs_ail_min(&(mp->m_ail.xa_ail));
+       spin_lock(&ailp->xa_lock);
+       lip = xfs_ail_min(ailp);
        if (lip == NULL) {
                lsn = (xfs_lsn_t)0;
        } else {
                lsn = lip->li_lsn;
        }
-       spin_unlock(&mp->m_ail_lock);
+       spin_unlock(&ailp->xa_lock);
 
        return lsn;
 }
@@ -78,23 +79,132 @@ xfs_trans_tail_ail(
  * the push is run asynchronously in a separate thread, so we return the tail
  * of the log right now instead of the tail after the push. This means we will
  * either continue right away, or we will sleep waiting on the async thread to
- * do it's work.
+ * do its work.
  *
  * We do this unlocked - we only need to know whether there is anything in the
  * AIL at the time we are called. We don't need to access the contents of
  * any of the objects, so the lock is not needed.
  */
 void
-xfs_trans_push_ail(
-       xfs_mount_t             *mp,
-       xfs_lsn_t               threshold_lsn)
+xfs_trans_ail_push(
+       struct xfs_ail  *ailp,
+       xfs_lsn_t       threshold_lsn)
 {
-       xfs_log_item_t          *lip;
+       xfs_log_item_t  *lip;
 
-       lip = xfs_ail_min(&mp->m_ail.xa_ail);
-       if (lip && !XFS_FORCED_SHUTDOWN(mp)) {
-               if (XFS_LSN_CMP(threshold_lsn, mp->m_ail.xa_target) > 0)
-                       xfsaild_wakeup(mp, threshold_lsn);
+       lip = xfs_ail_min(ailp);
+       if (lip && !XFS_FORCED_SHUTDOWN(ailp->xa_mount)) {
+               if (XFS_LSN_CMP(threshold_lsn, ailp->xa_target) > 0)
+                       xfsaild_wakeup(ailp, threshold_lsn);
+       }
+}
+
+/*
+ * AIL traversal cursor initialisation.
+ *
+ * The cursor keeps track of where our current traversal is up
+ * to by tracking the next ƣtem in the list for us. However, for
+ * this to be safe, removing an object from the AIL needs to invalidate
+ * any cursor that points to it. hence the traversal cursor needs to
+ * be linked to the struct xfs_ail so that deletion can search all the
+ * active cursors for invalidation.
+ *
+ * We don't link the push cursor because it is embedded in the struct
+ * xfs_ail and hence easily findable.
+ */
+STATIC void
+xfs_trans_ail_cursor_init(
+       struct xfs_ail          *ailp,
+       struct xfs_ail_cursor   *cur)
+{
+       cur->item = NULL;
+       if (cur == &ailp->xa_cursors)
+               return;
+
+       cur->next = ailp->xa_cursors.next;
+       ailp->xa_cursors.next = cur;
+}
+
+/*
+ * Set the cursor to the next item, because when we look
+ * up the cursor the current item may have been freed.
+ */
+STATIC void
+xfs_trans_ail_cursor_set(
+       struct xfs_ail          *ailp,
+       struct xfs_ail_cursor   *cur,
+       struct xfs_log_item     *lip)
+{
+       if (lip)
+               cur->item = xfs_ail_next(ailp, lip);
+}
+
+/*
+ * Get the next item in the traversal and advance the cursor.
+ * If the cursor was invalidated (inidicated by a lip of 1),
+ * restart the traversal.
+ */
+struct xfs_log_item *
+xfs_trans_ail_cursor_next(
+       struct xfs_ail          *ailp,
+       struct xfs_ail_cursor   *cur)
+{
+       struct xfs_log_item     *lip = cur->item;
+
+       if ((__psint_t)lip & 1)
+               lip = xfs_ail_min(ailp);
+       xfs_trans_ail_cursor_set(ailp, cur, lip);
+       return lip;
+}
+
+/*
+ * Now that the traversal is complete, we need to remove the cursor
+ * from the list of traversing cursors. Avoid removing the embedded
+ * push cursor, but use the fact it is always present to make the
+ * list deletion simple.
+ */
+void
+xfs_trans_ail_cursor_done(
+       struct xfs_ail          *ailp,
+       struct xfs_ail_cursor   *done)
+{
+       struct xfs_ail_cursor   *prev = NULL;
+       struct xfs_ail_cursor   *cur;
+
+       done->item = NULL;
+       if (done == &ailp->xa_cursors)
+               return;
+       prev = &ailp->xa_cursors;
+       for (cur = prev->next; cur; prev = cur, cur = prev->next) {
+               if (cur == done) {
+                       prev->next = cur->next;
+                       break;
+               }
+       }
+       ASSERT(cur);
+}
+
+/*
+ * Invalidate any cursor that is pointing to this item. This is
+ * called when an item is removed from the AIL. Any cursor pointing
+ * to this object is now invalid and the traversal needs to be
+ * terminated so it doesn't reference a freed object. We set the
+ * cursor item to a value of 1 so we can distinguish between an
+ * invalidation and the end of the list when getting the next item
+ * from the cursor.
+ */
+STATIC void
+xfs_trans_ail_cursor_clear(
+       struct xfs_ail          *ailp,
+       struct xfs_log_item     *lip)
+{
+       struct xfs_ail_cursor   *cur;
+
+       /* need to search all cursors */
+       for (cur = &ailp->xa_cursors; cur; cur = cur->next) {
+               if (cur->item == lip)
+                       cur->item = (struct xfs_log_item *)
+                                       ((__psint_t)cur->item | 1);
        }
 }
 
@@ -103,53 +213,58 @@ xfs_trans_push_ail(
  * Return the current tree generation number for use
  * in calls to xfs_trans_next_ail().
  */
-STATIC xfs_log_item_t *
-xfs_trans_first_push_ail(
-       xfs_mount_t     *mp,
-       int             *gen,
-       xfs_lsn_t       lsn)
+xfs_log_item_t *
+xfs_trans_ail_cursor_first(
+       struct xfs_ail          *ailp,
+       struct xfs_ail_cursor   *cur,
+       xfs_lsn_t               lsn)
 {
-       xfs_log_item_t  *lip;
+       xfs_log_item_t          *lip;
 
-       lip = xfs_ail_min(&(mp->m_ail.xa_ail));
-       *gen = (int)mp->m_ail.xa_gen;
+       xfs_trans_ail_cursor_init(ailp, cur);
+       lip = xfs_ail_min(ailp);
        if (lsn == 0)
-               return lip;
-
-       while (lip && (XFS_LSN_CMP(lip->li_lsn, lsn) < 0))
-               lip = lip->li_ail.ail_forw;
+               goto out;
 
+       list_for_each_entry(lip, &ailp->xa_ail, li_ail) {
+               if (XFS_LSN_CMP(lip->li_lsn, lsn) >= 0)
+                       goto out;
+       }
+       lip = NULL;
+out:
+       xfs_trans_ail_cursor_set(ailp, cur, lip);
        return lip;
 }
 
 /*
- * Function that does the work of pushing on the AIL
+ * xfsaild_push does the work of pushing on the AIL.  Returning a timeout of
+ * zero indicates that the caller should sleep until woken.
  */
 long
 xfsaild_push(
-       xfs_mount_t     *mp,
+       struct xfs_ail  *ailp,
        xfs_lsn_t       *last_lsn)
 {
-       long            tout = 1000; /* milliseconds */
+       long            tout = 0;
        xfs_lsn_t       last_pushed_lsn = *last_lsn;
-       xfs_lsn_t       target =  mp->m_ail.xa_target;
+       xfs_lsn_t       target =  ailp->xa_target;
        xfs_lsn_t       lsn;
        xfs_log_item_t  *lip;
-       int             gen;
-       int             restarts;
        int             flush_log, count, stuck;
+       xfs_mount_t     *mp = ailp->xa_mount;
+       struct xfs_ail_cursor   *cur = &ailp->xa_cursors;
 
-#define        XFS_TRANS_PUSH_AIL_RESTARTS     10
-
-       spin_lock(&mp->m_ail_lock);
-       lip = xfs_trans_first_push_ail(mp, &gen, *last_lsn);
+       spin_lock(&ailp->xa_lock);
+       xfs_trans_ail_cursor_init(ailp, cur);
+       lip = xfs_trans_ail_cursor_first(ailp, cur, *last_lsn);
        if (!lip || XFS_FORCED_SHUTDOWN(mp)) {
                /*
                 * AIL is empty or our push has reached the end.
                 */
-               spin_unlock(&mp->m_ail_lock);
-               last_pushed_lsn = 0;
-               goto out;
+               xfs_trans_ail_cursor_done(ailp, cur);
+               spin_unlock(&ailp->xa_lock);
+               *last_lsn = 0;
+               return tout;
        }
 
        XFS_STATS_INC(xs_push_ail);
@@ -165,9 +280,8 @@ xfsaild_push(
         * prevents use from spinning when we can't do anything or there is
         * lots of contention on the AIL lists.
         */
-       tout = 10;
        lsn = lip->li_lsn;
-       flush_log = stuck = count = restarts = 0;
+       flush_log = stuck = count = 0;
        while ((XFS_LSN_CMP(lip->li_lsn, target) < 0)) {
                int     lock_result;
                /*
@@ -182,7 +296,7 @@ xfsaild_push(
                 * skip to the next item in the list.
                 */
                lock_result = IOP_TRYLOCK(lip);
-               spin_unlock(&mp->m_ail_lock);
+               spin_unlock(&ailp->xa_lock);
                switch (lock_result) {
                case XFS_ITEM_SUCCESS:
                        XFS_STATS_INC(xs_push_ail_success);
@@ -219,7 +333,7 @@ xfsaild_push(
                        break;
                }
 
-               spin_lock(&mp->m_ail_lock);
+               spin_lock(&ailp->xa_lock);
                /* should we bother continuing? */
                if (XFS_FORCED_SHUTDOWN(mp))
                        break;
@@ -242,14 +356,13 @@ xfsaild_push(
                if (stuck > 100)
                        break;
 
-               lip = xfs_trans_next_ail(mp, lip, &gen, &restarts);
+               lip = xfs_trans_ail_cursor_next(ailp, cur);
                if (lip == NULL)
                        break;
-               if (restarts > XFS_TRANS_PUSH_AIL_RESTARTS)
-                       break;
                lsn = lip->li_lsn;
        }
-       spin_unlock(&mp->m_ail_lock);
+       xfs_trans_ail_cursor_done(ailp, cur);
+       spin_unlock(&ailp->xa_lock);
 
        if (flush_log) {
                /*
@@ -261,16 +374,18 @@ xfsaild_push(
                xfs_log_force(mp, (xfs_lsn_t)0, XFS_LOG_FORCE);
        }
 
-       /*
-        * We reached the target so wait a bit longer for I/O to complete and
-        * remove pushed items from the AIL before we start the next scan from
-        * the start of the AIL.
-        */
-       if ((XFS_LSN_CMP(lsn, target) >= 0)) {
-               tout += 20;
+       if (!count) {
+               /* We're past our target or empty, so idle */
                last_pushed_lsn = 0;
-       } else if ((restarts > XFS_TRANS_PUSH_AIL_RESTARTS) ||
-                  (count && ((stuck * 100) / count > 90))) {
+       } else if (XFS_LSN_CMP(lsn, target) >= 0) {
+               /*
+                * We reached the target so wait a bit longer for I/O to
+                * complete and remove pushed items from the AIL before we
+                * start the next scan from the start of the AIL.
+                */
+               tout = 50;
+               last_pushed_lsn = 0;
+       } else if ((stuck * 100) / count > 90) {
                /*
                 * Either there is a lot of contention on the AIL or we
                 * are stuck due to operations in progress. "Stuck" in this
@@ -280,12 +395,14 @@ xfsaild_push(
                 * Backoff a bit more to allow some I/O to complete before
                 * continuing from where we were.
                 */
-               tout += 10;
+               tout = 20;
+       } else {
+               /* more to do, but wait a short while before continuing */
+               tout = 10;
        }
-out:
        *last_lsn = last_pushed_lsn;
        return tout;
-}      /* xfsaild_push */
+}
 
 
 /*
@@ -298,7 +415,7 @@ out:
  */
 void
 xfs_trans_unlocked_item(
-       xfs_mount_t     *mp,
+       struct xfs_ail  *ailp,
        xfs_log_item_t  *lip)
 {
        xfs_log_item_t  *min_lip;
@@ -310,7 +427,7 @@ xfs_trans_unlocked_item(
         * over some potentially valid data.
         */
        if (!(lip->li_flags & XFS_LI_IN_AIL) ||
-           XFS_FORCED_SHUTDOWN(mp)) {
+           XFS_FORCED_SHUTDOWN(ailp->xa_mount)) {
                return;
        }
 
@@ -326,10 +443,10 @@ xfs_trans_unlocked_item(
         * the call to xfs_log_move_tail() doesn't do anything if there's
         * not enough free space to wake people up so we're safe calling it.
         */
-       min_lip = xfs_ail_min(&mp->m_ail.xa_ail);
+       min_lip = xfs_ail_min(ailp);
 
        if (min_lip == lip)
-               xfs_log_move_tail(mp, 1);
+               xfs_log_move_tail(ailp->xa_mount, 1);
 }      /* xfs_trans_unlocked_item */
 
 
@@ -342,43 +459,46 @@ xfs_trans_unlocked_item(
  * we move in the AIL is the minimum one, update the tail lsn in the
  * log manager.
  *
- * Increment the AIL's generation count to indicate that the tree
- * has changed.
- *
  * This function must be called with the AIL lock held.  The lock
  * is dropped before returning.
  */
 void
-xfs_trans_update_ail(
-       xfs_mount_t     *mp,
+xfs_trans_ail_update(
+       struct xfs_ail  *ailp,
        xfs_log_item_t  *lip,
-       xfs_lsn_t       lsn) __releases(mp->m_ail_lock)
+       xfs_lsn_t       lsn) __releases(ailp->xa_lock)
 {
-       xfs_ail_entry_t         *ailp;
-       xfs_log_item_t          *dlip=NULL;
+       xfs_log_item_t          *dlip = NULL;
        xfs_log_item_t          *mlip;  /* ptr to minimum lip */
+       xfs_lsn_t               tail_lsn;
 
-       ailp = &(mp->m_ail.xa_ail);
        mlip = xfs_ail_min(ailp);
 
        if (lip->li_flags & XFS_LI_IN_AIL) {
                dlip = xfs_ail_delete(ailp, lip);
                ASSERT(dlip == lip);
+               xfs_trans_ail_cursor_clear(ailp, dlip);
        } else {
                lip->li_flags |= XFS_LI_IN_AIL;
        }
 
        lip->li_lsn = lsn;
-
        xfs_ail_insert(ailp, lip);
-       mp->m_ail.xa_gen++;
 
        if (mlip == dlip) {
-               mlip = xfs_ail_min(&(mp->m_ail.xa_ail));
-               spin_unlock(&mp->m_ail_lock);
-               xfs_log_move_tail(mp, mlip->li_lsn);
+               mlip = xfs_ail_min(ailp);
+               /*
+                * It is not safe to access mlip after the AIL lock is
+                * dropped, so we must get a copy of li_lsn before we do
+                * so.  This is especially important on 32-bit platforms
+                * where accessing and updating 64-bit values like li_lsn
+                * is not atomic.
+                */
+               tail_lsn = mlip->li_lsn;
+               spin_unlock(&ailp->xa_lock);
+               xfs_log_move_tail(ailp->xa_mount, tail_lsn);
        } else {
-               spin_unlock(&mp->m_ail_lock);
+               spin_unlock(&ailp->xa_lock);
        }
 
 
@@ -400,31 +520,38 @@ xfs_trans_update_ail(
  * is dropped before returning.
  */
 void
-xfs_trans_delete_ail(
-       xfs_mount_t     *mp,
-       xfs_log_item_t  *lip) __releases(mp->m_ail_lock)
+xfs_trans_ail_delete(
+       struct xfs_ail  *ailp,
+       xfs_log_item_t  *lip) __releases(ailp->xa_lock)
 {
-       xfs_ail_entry_t         *ailp;
        xfs_log_item_t          *dlip;
        xfs_log_item_t          *mlip;
+       xfs_lsn_t               tail_lsn;
 
        if (lip->li_flags & XFS_LI_IN_AIL) {
-               ailp = &(mp->m_ail.xa_ail);
                mlip = xfs_ail_min(ailp);
                dlip = xfs_ail_delete(ailp, lip);
                ASSERT(dlip == lip);
+               xfs_trans_ail_cursor_clear(ailp, dlip);
 
 
                lip->li_flags &= ~XFS_LI_IN_AIL;
                lip->li_lsn = 0;
-               mp->m_ail.xa_gen++;
 
                if (mlip == dlip) {
-                       mlip = xfs_ail_min(&(mp->m_ail.xa_ail));
-                       spin_unlock(&mp->m_ail_lock);
-                       xfs_log_move_tail(mp, (mlip ? mlip->li_lsn : 0));
+                       mlip = xfs_ail_min(ailp);
+                       /*
+                        * It is not safe to access mlip after the AIL lock
+                        * is dropped, so we must get a copy of li_lsn
+                        * before we do so.  This is especially important
+                        * on 32-bit platforms where accessing and updating
+                        * 64-bit values like li_lsn is not atomic.
+                        */
+                       tail_lsn = mlip ? mlip->li_lsn : 0;
+                       spin_unlock(&ailp->xa_lock);
+                       xfs_log_move_tail(ailp->xa_mount, tail_lsn);
                } else {
-                       spin_unlock(&mp->m_ail_lock);
+                       spin_unlock(&ailp->xa_lock);
                }
        }
        else {
@@ -432,13 +559,13 @@ xfs_trans_delete_ail(
                 * If the file system is not being shutdown, we are in
                 * serious trouble if we get to this stage.
                 */
-               if (XFS_FORCED_SHUTDOWN(mp))
-                       spin_unlock(&mp->m_ail_lock);
-               else {
+               struct xfs_mount        *mp = ailp->xa_mount;
+
+               spin_unlock(&ailp->xa_lock);
+               if (!XFS_FORCED_SHUTDOWN(mp)) {
                        xfs_cmn_err(XFS_PTAG_AILDELETE, CE_ALERT, mp,
                "%s: attempting to delete a log item that is not in the AIL",
-                                       __FUNCTION__);
-                       spin_unlock(&mp->m_ail_lock);
+                                       __func__);
                        xfs_force_shutdown(mp, SHUTDOWN_CORRUPT_INCORE);
                }
        }
@@ -447,56 +574,6 @@ xfs_trans_delete_ail(
 
 
 /*
- * Return the item in the AIL with the smallest lsn.
- * Return the current tree generation number for use
- * in calls to xfs_trans_next_ail().
- */
-xfs_log_item_t *
-xfs_trans_first_ail(
-       xfs_mount_t     *mp,
-       int             *gen)
-{
-       xfs_log_item_t  *lip;
-
-       lip = xfs_ail_min(&(mp->m_ail.xa_ail));
-       *gen = (int)mp->m_ail.xa_gen;
-
-       return lip;
-}
-
-/*
- * If the generation count of the tree has not changed since the
- * caller last took something from the AIL, then return the elmt
- * in the tree which follows the one given.  If the count has changed,
- * then return the minimum elmt of the AIL and bump the restarts counter
- * if one is given.
- */
-xfs_log_item_t *
-xfs_trans_next_ail(
-       xfs_mount_t     *mp,
-       xfs_log_item_t  *lip,
-       int             *gen,
-       int             *restarts)
-{
-       xfs_log_item_t  *nlip;
-
-       ASSERT(mp && lip && gen);
-       if (mp->m_ail.xa_gen == *gen) {
-               nlip = xfs_ail_next(&(mp->m_ail.xa_ail), lip);
-       } else {
-               nlip = xfs_ail_min(&(mp->m_ail).xa_ail);
-               *gen = (int)mp->m_ail.xa_gen;
-               if (restarts != NULL) {
-                       XFS_STATS_INC(xs_push_ail_restarts);
-                       (*restarts)++;
-               }
-       }
-
-       return (nlip);
-}
-
-
-/*
  * The active item list (AIL) is a doubly linked list of log
  * items sorted by ascending lsn.  The base of the list is
  * a forw/back pointer pair embedded in the xfs mount structure.
@@ -514,16 +591,35 @@ int
 xfs_trans_ail_init(
        xfs_mount_t     *mp)
 {
-       mp->m_ail.xa_ail.ail_forw = (xfs_log_item_t*)&mp->m_ail.xa_ail;
-       mp->m_ail.xa_ail.ail_back = (xfs_log_item_t*)&mp->m_ail.xa_ail;
-       return xfsaild_start(mp);
+       struct xfs_ail  *ailp;
+       int             error;
+
+       ailp = kmem_zalloc(sizeof(struct xfs_ail), KM_MAYFAIL);
+       if (!ailp)
+               return ENOMEM;
+
+       ailp->xa_mount = mp;
+       INIT_LIST_HEAD(&ailp->xa_ail);
+       spin_lock_init(&ailp->xa_lock);
+       error = xfsaild_start(ailp);
+       if (error)
+               goto out_free_ailp;
+       mp->m_ail = ailp;
+       return 0;
+
+out_free_ailp:
+       kmem_free(ailp);
+       return error;
 }
 
 void
 xfs_trans_ail_destroy(
        xfs_mount_t     *mp)
 {
-       xfsaild_stop(mp);
+       struct xfs_ail  *ailp = mp->m_ail;
+
+       xfsaild_stop(ailp);
+       kmem_free(ailp);
 }
 
 /*
@@ -534,7 +630,7 @@ xfs_trans_ail_destroy(
  */
 STATIC void
 xfs_ail_insert(
-       xfs_ail_entry_t *base,
+       struct xfs_ail  *ailp,
        xfs_log_item_t  *lip)
 /* ARGSUSED */
 {
@@ -543,27 +639,22 @@ xfs_ail_insert(
        /*
         * If the list is empty, just insert the item.
         */
-       if (base->ail_back == (xfs_log_item_t*)base) {
-               base->ail_forw = lip;
-               base->ail_back = lip;
-               lip->li_ail.ail_forw = (xfs_log_item_t*)base;
-               lip->li_ail.ail_back = (xfs_log_item_t*)base;
+       if (list_empty(&ailp->xa_ail)) {
+               list_add(&lip->li_ail, &ailp->xa_ail);
                return;
        }
 
-       next_lip = base->ail_back;
-       while ((next_lip != (xfs_log_item_t*)base) &&
-              (XFS_LSN_CMP(next_lip->li_lsn, lip->li_lsn) > 0)) {
-               next_lip = next_lip->li_ail.ail_back;
+       list_for_each_entry_reverse(next_lip, &ailp->xa_ail, li_ail) {
+               if (XFS_LSN_CMP(next_lip->li_lsn, lip->li_lsn) <= 0)
+                       break;
        }
-       ASSERT((next_lip == (xfs_log_item_t*)base) ||
+
+       ASSERT((&next_lip->li_ail == &ailp->xa_ail) ||
               (XFS_LSN_CMP(next_lip->li_lsn, lip->li_lsn) <= 0));
-       lip->li_ail.ail_forw = next_lip->li_ail.ail_forw;
-       lip->li_ail.ail_back = next_lip;
-       next_lip->li_ail.ail_forw = lip;
-       lip->li_ail.ail_forw->li_ail.ail_back = lip;
 
-       xfs_ail_check(base, lip);
+       list_add(&lip->li_ail, &next_lip->li_ail);
+
+       xfs_ail_check(ailp, lip);
        return;
 }
 
@@ -573,15 +664,13 @@ xfs_ail_insert(
 /*ARGSUSED*/
 STATIC xfs_log_item_t *
 xfs_ail_delete(
-       xfs_ail_entry_t *base,
+       struct xfs_ail  *ailp,
        xfs_log_item_t  *lip)
 /* ARGSUSED */
 {
-       xfs_ail_check(base, lip);
-       lip->li_ail.ail_forw->li_ail.ail_back = lip->li_ail.ail_back;
-       lip->li_ail.ail_back->li_ail.ail_forw = lip->li_ail.ail_forw;
-       lip->li_ail.ail_forw = NULL;
-       lip->li_ail.ail_back = NULL;
+       xfs_ail_check(ailp, lip);
+
+       list_del(&lip->li_ail);
 
        return lip;
 }
@@ -592,14 +681,13 @@ xfs_ail_delete(
  */
 STATIC xfs_log_item_t *
 xfs_ail_min(
-       xfs_ail_entry_t *base)
+       struct xfs_ail  *ailp)
 /* ARGSUSED */
 {
-       register xfs_log_item_t *forw = base->ail_forw;
-       if (forw == (xfs_log_item_t*)base) {
+       if (list_empty(&ailp->xa_ail))
                return NULL;
-       }
-       return forw;
+
+       return list_first_entry(&ailp->xa_ail, xfs_log_item_t, li_ail);
 }
 
 /*
@@ -609,15 +697,14 @@ xfs_ail_min(
  */
 STATIC xfs_log_item_t *
 xfs_ail_next(
-       xfs_ail_entry_t *base,
+       struct xfs_ail  *ailp,
        xfs_log_item_t  *lip)
 /* ARGSUSED */
 {
-       if (lip->li_ail.ail_forw == (xfs_log_item_t*)base) {
+       if (lip->li_ail.next == &ailp->xa_ail)
                return NULL;
-       }
-       return lip->li_ail.ail_forw;
 
+       return list_first_entry(&lip->li_ail, xfs_log_item_t, li_ail);
 }
 
 #ifdef DEBUG
@@ -626,57 +713,40 @@ xfs_ail_next(
  */
 STATIC void
 xfs_ail_check(
-       xfs_ail_entry_t *base,
+       struct xfs_ail  *ailp,
        xfs_log_item_t  *lip)
 {
        xfs_log_item_t  *prev_lip;
 
-       prev_lip = base->ail_forw;
-       if (prev_lip == (xfs_log_item_t*)base) {
-               /*
-                * Make sure the pointers are correct when the list
-                * is empty.
-                */
-               ASSERT(base->ail_back == (xfs_log_item_t*)base);
+       if (list_empty(&ailp->xa_ail))
                return;
-       }
 
        /*
         * Check the next and previous entries are valid.
         */
        ASSERT((lip->li_flags & XFS_LI_IN_AIL) != 0);
-       prev_lip = lip->li_ail.ail_back;
-       if (prev_lip != (xfs_log_item_t*)base) {
-               ASSERT(prev_lip->li_ail.ail_forw == lip);
+       prev_lip = list_entry(lip->li_ail.prev, xfs_log_item_t, li_ail);
+       if (&prev_lip->li_ail != &ailp->xa_ail)
                ASSERT(XFS_LSN_CMP(prev_lip->li_lsn, lip->li_lsn) <= 0);
-       }
-       prev_lip = lip->li_ail.ail_forw;
-       if (prev_lip != (xfs_log_item_t*)base) {
-               ASSERT(prev_lip->li_ail.ail_back == lip);
+
+       prev_lip = list_entry(lip->li_ail.next, xfs_log_item_t, li_ail);
+       if (&prev_lip->li_ail != &ailp->xa_ail)
                ASSERT(XFS_LSN_CMP(prev_lip->li_lsn, lip->li_lsn) >= 0);
-       }
 
 
 #ifdef XFS_TRANS_DEBUG
        /*
-        * Walk the list checking forward and backward pointers,
-        * lsn ordering, and that every entry has the XFS_LI_IN_AIL
-        * flag set. This is really expensive, so only do it when
-        * specifically debugging the transaction subsystem.
+        * Walk the list checking lsn ordering, and that every entry has the
+        * XFS_LI_IN_AIL flag set. This is really expensive, so only do it
+        * when specifically debugging the transaction subsystem.
         */
-       prev_lip = (xfs_log_item_t*)base;
-       while (lip != (xfs_log_item_t*)base) {
-               if (prev_lip != (xfs_log_item_t*)base) {
-                       ASSERT(prev_lip->li_ail.ail_forw == lip);
+       prev_lip = list_entry(&ailp->xa_ail, xfs_log_item_t, li_ail);
+       list_for_each_entry(lip, &ailp->xa_ail, li_ail) {
+               if (&prev_lip->li_ail != &ailp->xa_ail)
                        ASSERT(XFS_LSN_CMP(prev_lip->li_lsn, lip->li_lsn) <= 0);
-               }
-               ASSERT(lip->li_ail.ail_back == prev_lip);
                ASSERT((lip->li_flags & XFS_LI_IN_AIL) != 0);
                prev_lip = lip;
-               lip = lip->li_ail.ail_forw;
        }
-       ASSERT(lip == (xfs_log_item_t*)base);
-       ASSERT(base->ail_back == prev_lip);
 #endif /* XFS_TRANS_DEBUG */
 }
 #endif /* DEBUG */