There two possible ways, print error and exit on too long names or
truncate the filename -- but it's impossible to do both in the same
code :-) It seems that code already assumes warning on long names, so
let's remove errx() and keep the behavior in dependence on -E.
Signed-off-by: Karel Zak <kzak@redhat.com>
Attempt to find and update all changes to command line options that has
not been reflected in bash-completion files in between versions v2.26 to
v2.27-rc1.
Signed-off-by: Sami Kerola <kerolasa@iki.fi>
- run the test with UID=0 only. The minix FS inode contains UID and
GID, so the final image checksum depends on user...
- always use subtest name for the images
Signed-off-by: Karel Zak <kzak@redhat.com>
The debian bug was fixed in commit 730ae9c899, and the rest are about
having test coverity for various file name lenghts and minix fs versions.
Signed-off-by: Sami Kerola <kerolasa@iki.fi>
Needed in order to create reproducable file systems image files, so that
out come of mkfs.minix can be checksum'ed.
Signed-off-by: Sami Kerola <kerolasa@iki.fi>
If a filesystem is mounted on top-level block device
with existing partitions, the mountpoint is not displayed
in the lsblk output.
This situation can happen by a configuration mistake
and lsblk could be used to detect such a mistake.
This patch allows searching for a mountpoint for all displayed
devices, not only for leaf nodes.
(It should be pretty cheap operation, mtab is parsed only once.)
For example: lsblk /dev/loop1
NAME MAJ:MIN RM SIZE RO TYPE MOUNTPOINT
loop1 7:1 0 128M 0 loop /mnt/tst
└─loop1p1 259:0 0 127M 0 loop
Signed-off-by: Milan Broz <gmazyland@gmail.com>
Use real table to device list, add missing new paragraph markers, remove
braces that cause reader to wonder if test is meant or not, squeeze error
code list, move references to bottom of the manual, and use mail or web
address macro when possible.
Signed-off-by: Sami Kerola <kerolasa@iki.fi>
Earlier user could define more blocks than device, or backing file for
loopback file system, had available. That lead to a system crash with
following commands;
fallocate --length 64KiB test-file
mkfs.minix -3 -i 842160 test-file 104882174
mkdir test-file.d
mount test-file test-file.d
cp /etc/service test-file.d
Killed
sudo umount test-file.d
The minix driver should probably not hang the whole kernel, but the least
that mkfs.minix ought to do is not to let users to get that condition
quite as easily.
Signed-off-by: Sami Kerola <kerolasa@iki.fi>
File name lenght and version input needs to be checked against each
other, which will determine what version of file system is in question.
Signed-off-by: Sami Kerola <kerolasa@iki.fi>
Checks about inodes vs block sizes does not add much robustness. Both
values are derived at compilation time from struct minix_inode size, and
they form full definition circle.
Bad block check for none-block devices should not be supressed, user
requested it so let him have it.
Check for st_rdev == 0x0300 || st_rdev == 0x0340 was unreachable.
Signed-off-by: Sami Kerola <kerolasa@iki.fi>
mkfs.minix misbehaves when attempting to create a large v2 or v3
filesystem. I finally traced it down to attempting to create too many
inodes so that the first zone is past 65535 blocks in. This obviously
doesn't work as the on-disk superblock says this is a 16 bit integer.
I wrote a patch that catches this, clamps to the absolute v2/v3 limit
(like it already does for v1), and sets the blocks per inode to a more
reasonable ratio when exceeding half a gigabyte. Having a half-gig
filesystem with most files being smaller than 3k isn't really reasonable.
I suppose if you don't want to adjust inode sizes automatically you could
take that part out, and it will just crab sooner.
Given the non-attention in the code, I suspect nobody ever had cause to
try such a big minix filesystem. Well I have my reasons involving some
deeply embedded work where ext2 would place too much strain on the
hardware.
Reviewed-by: Sami Kerola <kerolasa@iki.fi>
Signed-off-by: Joshua Hudson <joshudson@gmail.com>
The mnt_free_filesystems() relies on NULL as the last item in the
filesystems array. It's necessary to keep NULL there after failed
strdup() too, because we call mnt_free_filesystems() to deallocate
after error.
Signed-off-by: Karel Zak <kzak@redhat.com>
Karel Zak
Sami Kerola
Benno Schulenberg
Milan Broz
Joshua Hudson