From: Kiyoshi Ueda m->queue_io is set to block processing I/Os, and it needs to be kept while pg-init, which issues multiple path activations, is in progress. But m->queue is cleared when a path activation completes without error in pg_init_done(), even while other path activations are in progress. That may cause undesired -EIO on paths which are not complete activation. This patch fixes that by not clearing m->queue_io until all path activations complete. (Before the hardware handlers were moved into the SCSI layer, pg_init only used one path.) Signed-off-by: Kiyoshi Ueda Signed-off-by: Jun'ichi Nomura Signed-off-by: Alasdair G Kergon --- drivers/md/dm-mpath.c | 17 +++++++++++------ 1 file changed, 11 insertions(+), 6 deletions(-) Index: linux-2.6.33-rc8/drivers/md/dm-mpath.c =================================================================== --- linux-2.6.33-rc8.orig/drivers/md/dm-mpath.c +++ linux-2.6.33-rc8/drivers/md/dm-mpath.c @@ -1181,14 +1181,19 @@ static void pg_init_done(void *data, int m->current_pgpath = NULL; m->current_pg = NULL; } - } else if (!m->pg_init_required) { - m->queue_io = 0; + } else if (!m->pg_init_required) pg->bypassed = 0; - } - m->pg_init_in_progress--; - if (!m->pg_init_in_progress) - queue_work(kmultipathd, &m->process_queued_ios); + if (--m->pg_init_in_progress) + /* Activations of other paths are still on going */ + goto out; + + if (!m->pg_init_required) + m->queue_io = 0; + + queue_work(kmultipathd, &m->process_queued_ios); + +out: spin_unlock_irqrestore(&m->lock, flags); }