aboutsummaryrefslogtreecommitdiffstats
diff options
context:
space:
mode:
authorDavid Teigland <teigland@redhat.com>2006-11-01 10:31:48 -0500
committerSteven Whitehouse <swhiteho@redhat.com>2006-11-30 10:35:24 -0500
commit4b77f2c93d052adca8cc8690b9b5e7f8798f4ddd (patch)
treeb61c4923c355d36875bf878212dfc1b2f1f0f7ba
parent2cdc98aaf072d573df10c503d3b3b0b74e2a6d06 (diff)
[DLM] do full recover_locks barrier
Red Hat BZ 211914 The previous patch "[DLM] fix aborted recovery during node removal" was incomplete as discovered with further testing. It set the bit for the RS_LOCKS barrier but did not then wait for the barrier. This is often ok, but sometimes it will cause yet another recovery hang. If it's a new node that also has the lowest nodeid that skips the barrier wait, then it misses the important step of collecting and reporting the barrier status from the other nodes (which is the job of the low nodeid in the barrier wait routine). Signed-off-by: David Teigland <teigland@redhat.com> Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
-rw-r--r--fs/dlm/recoverd.c8
1 files changed, 7 insertions, 1 deletions
diff --git a/fs/dlm/recoverd.c b/fs/dlm/recoverd.c
index 6e4ee94ce7df..8bb895ffd90e 100644
--- a/fs/dlm/recoverd.c
+++ b/fs/dlm/recoverd.c
@@ -168,9 +168,15 @@ static int ls_recover(struct dlm_ls *ls, struct dlm_recover *rv)
168 /* 168 /*
169 * Other lockspace members may be going through the "neg" steps 169 * Other lockspace members may be going through the "neg" steps
170 * while also adding us to the lockspace, in which case they'll 170 * while also adding us to the lockspace, in which case they'll
171 * be looking for this status bit during dlm_recover_locks(). 171 * be doing the recover_locks (RS_LOCKS) barrier.
172 */ 172 */
173 dlm_set_recover_status(ls, DLM_RS_LOCKS); 173 dlm_set_recover_status(ls, DLM_RS_LOCKS);
174
175 error = dlm_recover_locks_wait(ls);
176 if (error) {
177 log_error(ls, "recover_locks_wait failed %d", error);
178 goto fail;
179 }
174 } 180 }
175 181
176 dlm_release_root_list(ls); 182 dlm_release_root_list(ls);