Red Hat Bugzilla – Bug 182529
Default targeted policy on FC4 blocks tune2fs
Last modified: 2007-11-30 17:11:24 EST
Description of problem: tune2fs fails to read the superblock of some (but not all) filesystems, when using LVM. Version-Release number of selected component (if applicable): tune2fs version tune2fs 1.38 (30-Jun-2005) How reproducible: always Steps to Reproduce: 1. see attached typescript 2. 3. Actual results: # tune2fs -l /dev/mapper/VolGroup00-LogVol03 tune2fs 1.38 (30-Jun-2005) ext2fs_check_if_mount: Permission denied while determining whether /dev/mapper/VolGroup00-LogVol03 is mounted. Expected results: tune2fs information Additional info: This one is wierd. As the typescript shows, some logical volumes work and others don't. Yet they were all created at the same time (at system install) using ext3. All the filesystems mount/umount without problem. I don't know why some work and not others. Work around: # setenforce 0
Created attachment 125076 [details] Typescript of session showing problem
In FC5 we are disabling the transition to fsadm_t from unconfined_t, so this would not be a problem there. fsadm_t is only used from init or from a confined domain. We are not intending up pushing any updates to FC4, so you could disable the transition to by changing the context on tune2fs to sbin_t chcon -t sbin_t /sbin/tune2fs