public inbox for cluster-cvs@sourceware.org
help / color / mirror / Atom feed
* Cluster Project branch, master, updated. cluster-2.99.05-2-g274ad92
@ 2008-06-24 21:57 teigland
  0 siblings, 0 replies; only message in thread
From: teigland @ 2008-06-24 21:57 UTC (permalink / raw)
  To: cluster-cvs, cluster-devel

This is an automated email from the git hooks/post-receive script. It was
generated because a ref change was pushed to the repository containing
the project "Cluster Project".

http://sources.redhat.com/git/gitweb.cgi?p=cluster.git;a=commitdiff;h=274ad923e82cf0567e129d1250c7f50ee7ca3b4b

The branch, master has been updated
       via  274ad923e82cf0567e129d1250c7f50ee7ca3b4b (commit)
      from  7fc8db7f157c928372f99412fc238a303c8f3b2d (commit)

Those revisions listed above that are new to this repository have
not appeared on any other notification email; so we list those
revisions in full, below.

- Log -----------------------------------------------------------------
commit 274ad923e82cf0567e129d1250c7f50ee7ca3b4b
Author: David Teigland <teigland@redhat.com>
Date:   Tue Jun 24 16:52:12 2008 -0500

    gfs_controld: basic fixes
    
    Fix leave/unmount; weren't calling the function.
    Fix match_change for messages from new nodes that don't contain
    info for all members.
    
    Signed-off-by: David Teigland <teigland@redhat.com>

-----------------------------------------------------------------------

Summary of changes:
 group/gfs_controld/cpg-new.c |   41 +++++++++++++++++++++++++++++++----------
 group/gfs_controld/main.c    |    5 +++--
 2 files changed, 34 insertions(+), 12 deletions(-)

diff --git a/group/gfs_controld/cpg-new.c b/group/gfs_controld/cpg-new.c
index f765e48..d43d53b 100644
--- a/group/gfs_controld/cpg-new.c
+++ b/group/gfs_controld/cpg-new.c
@@ -780,21 +780,30 @@ static int match_change(struct mountgroup *mg, struct change *cg,
 	uint32_t seq = hd->msgdata;
 	int i, members_mismatch;
 
-	/* We can ignore messages if we're not in the list of members.  The one
-	   known time this will happen is after we've joined the cpg, we can
-	   get messages for changes prior to the change in which we're added. */
-
 	ids = (struct id_info *)((char *)hd +
 				 sizeof(struct gfs_header) +
 				 mi->mg_info_size);
 
-	id = get_id_struct(ids, mi->id_info_count, mi->id_info_size,
-			   our_nodeid);
+	if (!mi->started_count) {
+		if (mi->id_info_count != 1) {
+			log_error("match_change fail %d:%u id_count %d exp 1",
+				  hd->nodeid, seq, mi->id_info_count);
+			return 0;
+		}
+	} else {
+		/* We can ignore messages if we're not in the list of members.
+		   The one known time this will happen is after we've joined
+		   the cpg, we can get messages for changes prior to the change
+		   in which we're added. */
 
-	if (!id || !(id->flags & IDI_NODEID_IS_MEMBER)) {
-		log_group(mg, "match_change fail %d:%u we are not in members",
-			  hd->nodeid, seq);
-		return 0;
+		id = get_id_struct(ids, mi->id_info_count, mi->id_info_size,
+				   our_nodeid);
+
+		if (!id || !(id->flags & IDI_NODEID_IS_MEMBER)) {
+			log_group(mg, "match_change fail %d:%u we are not in "
+				  "members", hd->nodeid, seq);
+			return 0;
+		}
 	}
 
 	memb = find_memb(cg, hd->nodeid);
@@ -1870,6 +1879,18 @@ void process_recovery_uevent(char *table)
 	}
 
 	if (!mg->first_recovery_needed) {
+		if (!mg->local_recovery_busy) {
+			/* we expect a recovery_done uevent for our own journal
+			   when we mount; shouldn't happen otherwise */
+
+			if (jid == mg->our_jid)
+				return;
+
+			log_error("process_recovery_uevent jid %d unexpected",
+				  jid);
+			return;
+		}
+
 		mg->local_recovery_busy = 0;
 
 		if (mg->local_recovery_jid != jid) {
diff --git a/group/gfs_controld/main.c b/group/gfs_controld/main.c
index 8fed0d2..60bd9b7 100644
--- a/group/gfs_controld/main.c
+++ b/group/gfs_controld/main.c
@@ -297,8 +297,7 @@ static void process_uevent(int ci)
 		if (lock_module)
 			return;
 
-		if (group_mode == GROUP_LIBGROUP)
-			do_leave(argv[3], 0);
+		do_leave(argv[3], 0);
 
 	} else if (!strcmp(act, "change@")) {
 		if (!lock_module)
@@ -315,6 +314,8 @@ static void process_uevent(int ci)
 
 		if (group_mode == GROUP_LIBGROUP)
 			do_withdraw_old(argv[3]);
+		else
+			log_error("TODO withdraw for libcpg");
 
 	} else {
 		if (!lock_module)


hooks/post-receive
--
Cluster Project


^ permalink raw reply	[flat|nested] only message in thread

only message in thread, other threads:[~2008-06-24 21:57 UTC | newest]

Thread overview: (only message) (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2008-06-24 21:57 Cluster Project branch, master, updated. cluster-2.99.05-2-g274ad92 teigland

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).