http://bugzilla.kernel.org/show_bug.cgi?id=14601 Summary: Wrong handling of META_BG when getting number of blocks for group descriptors Product: File System Version: 2.5 Platform: All OS/Version: Linux Tree: Mainline Status: NEW Severity: normal Priority: P1 Component: ext4 AssignedTo: fs_ext4@xxxxxxxxxxxxxxxxxxxx ReportedBy: DamienDGU@xxxxxxxxxxx Regression: No Created an attachment (id=23778) --> (http://bugzilla.kernel.org/attachment.cgi?id=23778) ext3/balloc.c and ext4/balloc.c change to count number of descriptors blocks When computing number of blocks for group descriptors for a given group, the total number of blocks for group descriptors is returned in case META_BG is enabled and group is inside the first groups (such as block of its descriptor is below s_first_meta_bg). I think it should be limited to s_first_meta_bg. This behaviour is incoherent with e2fsprogs behaviour (see http://marc.info/?l=linux-ext4&m=125794126219545&w=2) and can lead to corruption in case performing e2fsck on a filesystem where an uninitialized block bitmap was initialized by kernel. I attach a proposed change, but: - I did not compiled, nor tried it. - it is for ext3 and ext4. - I am not sure it is enough: there is some other location where s_gdb_count is used and where perhaps ext4_bg_num_gdb should be used instead as in ext4_setup_system_zone (I do not understand the behaviour here, it seems to not take META_BG flag into account at all) or ex4_iget. -- Configure bugmail: http://bugzilla.kernel.org/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are watching the assignee of the bug. -- To unsubscribe from this list: send the line "unsubscribe linux-ext4" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
