Skip to content

Commit

Permalink
xfs: fix dquot shaker deadlock
Browse files Browse the repository at this point in the history
Commit 368e136 ("xfs: remove duplicate code from dquot reclaim") fails
to unlock the dquot freelist when the number of loop restarts is
exceeded in xfs_qm_dqreclaim_one(). This causes hangs in memory
reclaim.

Rework the loop control logic into an unwind stack that all the
different cases jump into. This means there is only one set of code
that processes the loop exit criteria, and simplifies the unlocking
of all the items from different points in the loop. It also fixes a
double increment of the restart counter from the qi_dqlist_lock
case.

Reported-by: Malcolm Scott <lkml@malc.org.uk>
Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Alex Elder <aelder@sgi.com>
  • Loading branch information
Dave Chinner authored and Alex Elder committed Jan 28, 2011
1 parent c6f990d commit 0fbca4d
Showing 1 changed file with 21 additions and 25 deletions.
46 changes: 21 additions & 25 deletions fs/xfs/quota/xfs_qm.c
Original file line number Diff line number Diff line change
Expand Up @@ -1863,12 +1863,14 @@ xfs_qm_dqreclaim_one(void)
xfs_dquot_t *dqpout;
xfs_dquot_t *dqp;
int restarts;
int startagain;

restarts = 0;
dqpout = NULL;

/* lockorder: hashchainlock, freelistlock, mplistlock, dqlock, dqflock */
startagain:
again:
startagain = 0;
mutex_lock(&xfs_Gqm->qm_dqfrlist_lock);

list_for_each_entry(dqp, &xfs_Gqm->qm_dqfrlist, q_freelist) {
Expand All @@ -1885,13 +1887,10 @@ xfs_qm_dqreclaim_one(void)
ASSERT(! (dqp->dq_flags & XFS_DQ_INACTIVE));

trace_xfs_dqreclaim_want(dqp);

xfs_dqunlock(dqp);
mutex_unlock(&xfs_Gqm->qm_dqfrlist_lock);
if (++restarts >= XFS_QM_RECLAIM_MAX_RESTARTS)
return NULL;
XQM_STATS_INC(xqmstats.xs_qm_dqwants);
goto startagain;
restarts++;
startagain = 1;
goto dqunlock;
}

/*
Expand All @@ -1906,23 +1905,20 @@ xfs_qm_dqreclaim_one(void)
ASSERT(list_empty(&dqp->q_mplist));
list_del_init(&dqp->q_freelist);
xfs_Gqm->qm_dqfrlist_cnt--;
xfs_dqunlock(dqp);
dqpout = dqp;
XQM_STATS_INC(xqmstats.xs_qm_dqinact_reclaims);
break;
goto dqunlock;
}

ASSERT(dqp->q_hash);
ASSERT(!list_empty(&dqp->q_mplist));

/*
* Try to grab the flush lock. If this dquot is in the process of
* getting flushed to disk, we don't want to reclaim it.
* Try to grab the flush lock. If this dquot is in the process
* of getting flushed to disk, we don't want to reclaim it.
*/
if (!xfs_dqflock_nowait(dqp)) {
xfs_dqunlock(dqp);
continue;
}
if (!xfs_dqflock_nowait(dqp))
goto dqunlock;

/*
* We have the flush lock so we know that this is not in the
Expand All @@ -1944,8 +1940,7 @@ xfs_qm_dqreclaim_one(void)
xfs_fs_cmn_err(CE_WARN, mp,
"xfs_qm_dqreclaim: dquot %p flush failed", dqp);
}
xfs_dqunlock(dqp); /* dqflush unlocks dqflock */
continue;
goto dqunlock;
}

/*
Expand All @@ -1967,13 +1962,8 @@ xfs_qm_dqreclaim_one(void)
*/
if (!mutex_trylock(&mp->m_quotainfo->qi_dqlist_lock)) {
restarts++;
mutex_unlock(&dqp->q_hash->qh_lock);
xfs_dqfunlock(dqp);
xfs_dqunlock(dqp);
mutex_unlock(&xfs_Gqm->qm_dqfrlist_lock);
if (restarts++ >= XFS_QM_RECLAIM_MAX_RESTARTS)
return NULL;
goto startagain;
startagain = 1;
goto qhunlock;
}

ASSERT(dqp->q_nrefs == 0);
Expand All @@ -1986,14 +1976,20 @@ xfs_qm_dqreclaim_one(void)
xfs_Gqm->qm_dqfrlist_cnt--;
dqpout = dqp;
mutex_unlock(&mp->m_quotainfo->qi_dqlist_lock);
qhunlock:
mutex_unlock(&dqp->q_hash->qh_lock);
dqfunlock:
xfs_dqfunlock(dqp);
dqunlock:
xfs_dqunlock(dqp);
if (dqpout)
break;
if (restarts >= XFS_QM_RECLAIM_MAX_RESTARTS)
return NULL;
break;
if (startagain) {
mutex_unlock(&xfs_Gqm->qm_dqfrlist_lock);
goto again;
}
}
mutex_unlock(&xfs_Gqm->qm_dqfrlist_lock);
return dqpout;
Expand Down

0 comments on commit 0fbca4d

Please sign in to comment.