From: Fan Yong <fan.yong@intel.com>
Date: Fri, 17 Jun 2016 03:20:39 +0000 (+0800)
Subject: LU-6635 lfsck: block replacing the OST-object for test
X-Git-Tag: 2.8.60~52
X-Git-Url: https://git.whamcloud.com/?p=fs%2Flustre-release.git;a=commitdiff_plain;h=7a814e94e065551ab79e2ba75df9626e4940efc5;hp=51f1b46dab6c2cc16f71d3f3be483f530b5e4001

LU-6635 lfsck: block replacing the OST-object for test

For sanity-lfsck test_18e, sometimes, before the client wirte
happened, the layout LFSCK has already gone into the phase2
and replaced the new created OST-object with the old orphan
OST-object, then cause the subsequent check failure.

To resolve such trouble, we will hold the layout LFSCK when
replacing the new created OST-object until the client write
happened.

Signed-off-by: Fan Yong <fan.yong@intel.com>
Change-Id: Ia5be2cd2ea920fa02bb281ca5c59d8d252f3fd7b
Reviewed-on: http://review.whamcloud.com/18146
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Reviewed-by: Lai Siyao <lai.siyao@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
---

diff --git a/lustre/lfsck/lfsck_layout.c b/lustre/lfsck/lfsck_layout.c
index d1cff12..56890b6 100644
--- a/lustre/lfsck/lfsck_layout.c
+++ b/lustre/lfsck/lfsck_layout.c
@@ -2089,6 +2089,11 @@ static int lfsck_layout_conflict_create(const struct lu_env *env,
 	int			  rc		= 0;
 	ENTRY;
 
+	while (CFS_FAIL_TIMEOUT(OBD_FAIL_LFSCK_DELAY3, cfs_fail_val)) {
+		if (unlikely(!thread_is_running(&com->lc_lfsck->li_thread)))
+			RETURN(0);
+	}
+
 	ostid_le_to_cpu(&slot->l_ost_oi, oi);
 	rc = ostid_to_fid(cfid2, oi, ost_idx2);
 	if (rc != 0)