kpartx -d treats any map that has a "linear" mapping into a
device as first target as a partition of this device.
This is wrong, because linear mappings may consist of
several pieces, combining multiple devices into one
(LVM logical volumes are an example of such a
mapping). Partitions have to be single-target mappings.

Fix this by returning an error in dm_type() if a map with
multiple targets is encountered. The "type" of a map with
two or more targets is a difficult concept, anyway.

test case:
truncate -s 1G test0 test1
losetup /dev/loop0 test0
losetup /dev/loop1 test1
dmsetup create <<EOF
0 1000000 linear 7:0 4096
1000000 1000000 linear 7:1 4096
EOF
kpartx -d /dev/loop0 -v

The map "join" should NOT be deleted.

Signed-off-by: Martin Wilck <[email protected]>
---
 kpartx/devmapper.c | 11 ++++++-----
 1 file changed, 6 insertions(+), 5 deletions(-)

diff --git a/kpartx/devmapper.c b/kpartx/devmapper.c
index d6ccd000..5380b2e9 100644
--- a/kpartx/devmapper.c
+++ b/kpartx/devmapper.c
@@ -390,10 +390,11 @@ dm_type(const char * name, char * type)
                goto out;
 
        /* Fetch 1st target */
-       dm_get_next_target(dmt, NULL, &start, &length,
-                          &target_type, &params);
-
-       if (!target_type)
+       if (dm_get_next_target(dmt, NULL, &start, &length,
+                              &target_type, &params) != NULL)
+               /* more than one target */
+               r = -1;
+       else if (!target_type)
                r = -1;
        else if (!strcmp(target_type, type))
                r = 1;
@@ -494,7 +495,7 @@ do_foreach_partmaps (const char * mapname, const char *uuid,
                /*
                 * skip if devmap target is not "linear"
                 */
-               if (!dm_type(names->name, "linear")) {
+               if (dm_type(names->name, "linear") != 1) {
                        if (rd->verbose)
                                printf("%s: is not a linear target. Not 
removing\n",
                                       names->name);
-- 
2.12.2

--
dm-devel mailing list
[email protected]
https://www.redhat.com/mailman/listinfo/dm-devel

Reply via email to