From: Nathaniel Clark <nathaniel.l.clark@intel.com>
Date: Thu, 10 Aug 2017 14:20:04 +0000 (-0400)
Subject: LU-9799 mount: Call read_ldd with initialized mount type
X-Git-Tag: 2.10.52~13
X-Git-Url: https://git.whamcloud.com/?a=commitdiff_plain;h=0108281c65545df169faaa0ce0690fb021680643;p=fs%2Flustre-release.git

LU-9799 mount: Call read_ldd with initialized mount type

When re-reading the ldd when doing relabel, ensure the correct mount
type is used for the read. Otherwise ldiskfs complains:

   mount.lustre FATAL: unhandled/unloaded fs type 0 'ext3'

and ZFS complains:

   e2label: No such file or directory while trying to open MGS/MGT
   Couldn't find valid filesystem superblock.

Signed-off-by: Nathaniel Clark <nathaniel.l.clark@intel.com>
Change-Id: Ife53cff948d545c306e99e4b023989245a1ac3f7
Reviewed-on: https://review.whamcloud.com/28456
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Fan Yong <fan.yong@intel.com>
Reviewed-by: Hongchao Zhang <hongchao.zhang@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
---

diff --git a/lustre/utils/mount_lustre.c b/lustre/utils/mount_lustre.c
index 4c6e597..de3833f 100644
--- a/lustre/utils/mount_lustre.c
+++ b/lustre/utils/mount_lustre.c
@@ -720,6 +720,7 @@ static void label_lustre(struct mount_opts *mop)
 		/* device label could be changed after journal recovery,
 		 * it should also be relabeled for mount has succeeded. */
 		memset(&ldd, 0, sizeof(ldd));
+		ldd.ldd_mount_type = mop->mo_ldd.ldd_mount_type;
 		rc = osd_read_ldd(mop->mo_source, &ldd);
 		if (rc == 0) {
 			rc = strlen(ldd.ldd_svname);