From 4875647a08e35f77274838d97ca8fa44158d50e2 Mon Sep 17 00:00:00 2001 From: David Teigland Date: Thu, 26 Apr 2012 15:54:29 -0500 Subject: dlm: fixes for nodir mode The "nodir" mode (statically assign master nodes instead of using the resource directory) has always been highly experimental, and never seriously used. This commit fixes a number of problems, making nodir much more usable. - Major change to recovery: recover all locks and restart all in-progress operations after recovery. In some cases it's not possible to know which in-progess locks to recover, so recover all. (Most require recovery in nodir mode anyway since rehashing changes most master nodes.) - Change the way nodir mode is enabled, from a command line mount arg passed through gfs2, into a sysfs file managed by dlm_controld, consistent with the other config settings. - Allow recovering MSTCPY locks on an rsb that has not yet been turned into a master copy. - Ignore RCOM_LOCK and RCOM_LOCK_REPLY recovery messages from a previous, aborted recovery cycle. Base this on the local recovery status not being in the state where any nodes should be sending LOCK messages for the current recovery cycle. - Hold rsb lock around dlm_purge_mstcpy_locks() because it may run concurrently with dlm_recover_master_copy(). - Maintain highbast on process-copy lkb's (in addition to the master as is usual), because the lkb can switch back and forth between being a master and being a process copy as the master node changes in recovery. - When recovering MSTCPY locks, flag rsb's that have non-empty convert or waiting queues for granting at the end of recovery. (Rename flag from LOCKS_PURGED to RECOVER_GRANT and similar for the recovery function, because it's not only resources with purged locks that need grant a grant attempt.) - Replace a couple of unnecessary assertion panics with error messages. Signed-off-by: David Teigland --- fs/dlm/requestqueue.c | 39 ++++++++++----------------------------- 1 file changed, 10 insertions(+), 29 deletions(-) (limited to 'fs/dlm/requestqueue.c') diff --git a/fs/dlm/requestqueue.c b/fs/dlm/requestqueue.c index d3191bf03a68..1695f1b0dd45 100644 --- a/fs/dlm/requestqueue.c +++ b/fs/dlm/requestqueue.c @@ -65,6 +65,7 @@ void dlm_add_requestqueue(struct dlm_ls *ls, int nodeid, struct dlm_message *ms) int dlm_process_requestqueue(struct dlm_ls *ls) { struct rq_entry *e; + struct dlm_message *ms; int error = 0; mutex_lock(&ls->ls_requestqueue_mutex); @@ -78,6 +79,14 @@ int dlm_process_requestqueue(struct dlm_ls *ls) e = list_entry(ls->ls_requestqueue.next, struct rq_entry, list); mutex_unlock(&ls->ls_requestqueue_mutex); + ms = &e->request; + + log_limit(ls, "dlm_process_requestqueue msg %d from %d " + "lkid %x remid %x result %d seq %u", + ms->m_type, ms->m_header.h_nodeid, + ms->m_lkid, ms->m_remid, ms->m_result, + e->recover_seq); + dlm_receive_message_saved(ls, &e->request, e->recover_seq); mutex_lock(&ls->ls_requestqueue_mutex); @@ -140,35 +149,7 @@ static int purge_request(struct dlm_ls *ls, struct dlm_message *ms, int nodeid) if (!dlm_no_directory(ls)) return 0; - /* with no directory, the master is likely to change as a part of - recovery; requests to/from the defunct master need to be purged */ - - switch (type) { - case DLM_MSG_REQUEST: - case DLM_MSG_CONVERT: - case DLM_MSG_UNLOCK: - case DLM_MSG_CANCEL: - /* we're no longer the master of this resource, the sender - will resend to the new master (see waiter_needs_recovery) */ - - if (dlm_hash2nodeid(ls, ms->m_hash) != dlm_our_nodeid()) - return 1; - break; - - case DLM_MSG_REQUEST_REPLY: - case DLM_MSG_CONVERT_REPLY: - case DLM_MSG_UNLOCK_REPLY: - case DLM_MSG_CANCEL_REPLY: - case DLM_MSG_GRANT: - /* this reply is from the former master of the resource, - we'll resend to the new master if needed */ - - if (dlm_hash2nodeid(ls, ms->m_hash) != nodeid) - return 1; - break; - } - - return 0; + return 1; } void dlm_purge_requestqueue(struct dlm_ls *ls) -- cgit v1.2.1