The error path was a bit unclear on the semantics. The seq argument
would get freed but not the name argument. Then the caller frees
them both when the function returns an error.
This fixes the problem my making the function not touch the arguments
unless we're going to return success. Also improve the description
for this function to make it more obvious what the callers should expect.
Addresses: CID#77487 (on scan.coverity.com for JIghtuse/util-linux) Signed-off-by: Andreas Henriksson <andreas@fatal.se>
Direct leak of 8 byte(s) in 1 object(s) allocated from:
#0 0x49d12b in __interceptor_malloc (/home/src/util-linux/.libs/lt-mkswap+0x49d12b)
#1 0x7faf2a5069c9 in __GI___strdup (/usr/lib/libc.so.6+0x819c9)
#2 0xffff96e7e33 (<unknown module>)
SUMMARY: AddressSanitizer: 8 byte(s) leaked in 1 allocation(s).
And another one that valgrind found.
==6316== 8 bytes in 1 blocks are definitely lost in loss record 1 of 1
==6316== at 0x4C29F90: malloc (in /usr/lib/valgrind/vgpreload_memcheck-amd64-linux.so)
==6316== by 0x5E3F9C9: strdup (in /usr/lib/libc-2.20.so)
==6316== by 0x43A25F: size_to_human_string (strutils.c:495)
==6316== by 0x42B35C: main (mkswap.c:488)
Sami Kerola [Sun, 9 Nov 2014 15:26:05 +0000 (15:26 +0000)]
libmount: fix memory overflow [AddressSanitizer]
==10918==ERROR: AddressSanitizer: stack-buffer-overflow on address 0x7fffd795b680 at pc 0x0000004447c6 bp 0x7fffd795b3e0 sp 0x7fffd795ab78
WRITE of size 129 at 0x7fffd795b680 thread T0
#0 0x4447c5 in scanf_common(void*, int, bool, char const*, __va_list_tag*) (/home/src/util-linux/.libs/lt-mount+0x4447c5)
#1 0x445892 in sscanf (/home/src/util-linux/.libs/lt-mount+0x445892)
#2 0x7fe78709a3d3 in get_filesystems /home/src/util-linux/libmount/src/utils.c:581:7
#3 0x7fe78709a1ba in mnt_get_filesystems /home/src/util-linux/libmount/src/utils.c:622:7
#4 0x7fe7870aa78f in do_mount_by_pattern /home/src/util-linux/libmount/src/context_mount.c:833:7
#5 0x7fe7870a9534 in mnt_context_do_mount /home/src/util-linux/libmount/src/context_mount.c:951:9
#6 0x7fe7870aab2b in mnt_context_mount /home/src/util-linux/libmount/src/context_mount.c:1051:8
#7 0x4ba9f5 in main /home/src/util-linux/sys-utils/mount.c:1107:7
#8 0x7fe785caa03f in __libc_start_main (/usr/lib/libc.so.6+0x2003f)
#9 0x4b9f9c in _start (/home/src/util-linux/.libs/lt-mount+0x4b9f9c)
Signed-off-by: Sami Kerola <kerolasa@iki.fi> Signed-off-by: Karel Zak <kzak@redhat.com>
Sami Kerola [Sun, 2 Nov 2014 22:06:15 +0000 (22:06 +0000)]
build-sys: make scanf_cv_alloc_modifier to work [LeakSanitizer]
The tests failed with following message in config.log
ERROR: LeakSanitizer: detected memory leaks
Direct leak of 2 byte(s) in 1 object(s) allocated from:
#0 0x49a40e in realloc (/home/src/util-linux/conftest+0x49a40e)
#1 0x7fbe48633e69 in __GI__IO_vfscanf (/usr/lib/libc.so.6+0x56e69)
#2 0x7fbe48649786 in _IO_vsscanf (/usr/lib/libc.so.6+0x6c786)
which knocked out libmount from build, and commands depending on it.
The reason this change makes sense is that AddressSanitizer seems like a
good addition to set of tools that util-linux package can use, when and
if needed.
Reference: https://code.google.com/p/address-sanitizer/wiki/AddressSanitizer Reviewed-by: Mike Frysinger <vapier@gentoo.org> Signed-off-by: Sami Kerola <kerolasa@iki.fi>
there is a theoretical buffer overflow possible in the hypervisor
parsing code of lscpu. It would require a proc entry to return way more
than expected so it's no high priority. But better be safe than sorry.
At first I thought about switching to fgets but there is another
code file that adds a format specifier. The diff is less intrusive
that way, too.
Boris Egorov [Fri, 14 Nov 2014 05:27:16 +0000 (11:27 +0600)]
libfdisk: fix get_partition_unused_primary()
Was:
Mentioned function returns -1 if adding of primary partition is
impossible. Caller treats this value as size_t (res variable) and then
compares it for negative values, totally ignoring errors.
Becomes:
Now function takes address to variable and fills it with partition
number. Caller treats return value as int and use it appropriately.
[kzak@redhat.com: - don't mix return code and partno at all]
Signed-off-by: Boris Egorov <egorov@linux.com> Signed-off-by: Karel Zak <kzak@redhat.com>
Karel Zak [Thu, 13 Nov 2014 10:19:58 +0000 (11:19 +0100)]
cfdisk: remove [Load] from menu
This functionality will be rarely used, so it does not make sense to
waste screen space with this menu item. It's enough to provide this
functionality in "Select label type" dialog (cfdisk --zero and then
'L' command).
Karel Zak [Wed, 12 Nov 2014 10:15:06 +0000 (11:15 +0100)]
fdisk: add support for sfdisk scripts
New commands 'I' and 'O' allows to read and write sfdisk compatible
scripts by fdisk. It means that you can save your work (partition
table) and later use it (in fdisk, sfdisk or cfdisk) to create a new
partition table.
Karel Zak [Mon, 10 Nov 2014 13:01:09 +0000 (14:01 +0100)]
libfdisk: (gpt) don't offer first sector before the first partion
The GPT first usable LBA is usually aligned to grain (1MiB), but for small
(<=4MiB) devices we strictly follow sector sizes.
In this case there is a small space in front of the aligned begin of
the first partition. This useless space should not be offered for the
next partitions.
Device Start End Size Type
/dev/sdb1 40 2087 1M Linux filesystem
Command (m for help): n
Partition number (2-128, default 2):
First sector (34-8158, default 2088):
^^
first usable LBA is 34, but first aligned (recommended) LBA is 40, we
use it for the first partition. All this is correct, but the space
before the first partition should be ignored. Fixed version:
Command (m for help): n
Partition number (2-128, default 2):
First sector (2088-8158, default 2088):
^^^^
Note this problem does not exist for "normal" (large) devices where
first usable LBA is aligned to grain.
Reported-by: Boaz Harrosh <boaz@plexistor.com> Signed-off-by: Karel Zak <kzak@redhat.com>
Karel Zak [Mon, 10 Nov 2014 11:22:55 +0000 (12:22 +0100)]
libfdisk: (gpt) make mode more readable
* use check_ prefix for checking functions
* use plural in function names where we work with more partitions than one
* always use "ents" when when we work with gpt entries array
Sami Kerola [Sun, 2 Nov 2014 20:26:31 +0000 (20:26 +0000)]
mkswap: various minor improvement
Use correct data type in page_bad(), and add information to error message
how many bad pages were seen.
In check_blocks() move initialization to variable introduction, fix typo,
and avoid memset() when array initializer can do the job.
In main() use correct initializer for pointer. Move swap file specific
actions to same location, and add warning when request too --check is
dismissed by command. At the end remove momory leak.
Sami Kerola [Sun, 2 Nov 2014 20:26:25 +0000 (20:26 +0000)]
mkswap: remove system architecture specific max swap size checks
Since kernel version 2.3.4 (June 1999) all architectures has used
uint32_t as maximum number or pages in a swap device or file, there is no
longer need to support systems earlier than that.
Karel Zak [Thu, 6 Nov 2014 11:50:27 +0000 (12:50 +0100)]
switch_root: improve statfs->f_type portability
__SWORD_TYPE is not available everywhere, for example it's not defined
by musl libc. It also seems that __SWORD_TYPE is not used for f_type
on some architectures (s390x).
Reported-by: Natanael Copa <ncopa@alpinelinux.org> Signed-off-by: Karel Zak <kzak@redhat.com>
Karel Zak [Mon, 3 Nov 2014 11:51:17 +0000 (12:51 +0100)]
include/debug: improve and cleanup
* use ul_debug_ prefix for all routines
* support <NAME>_DEBUG=all also for programs without debug mask names
(so we can avoid 0xffff mask in man pages)
* add function to print debug help
Mike Frysinger [Wed, 29 Oct 2014 04:19:03 +0000 (00:19 -0400)]
libmount: improve kernel command line parsing
The current command line parser will stop at the first occurrence of an
option, however the kernel does the opposite. So if you have:
root=/dev/sda1 root=/dev/sda2
When you look for "root", the kernel will use /dev/sda2, but util-linux
uses /dev/sda1.
Further, if args are passed to custom init programs, the parser will
pick those up as kernel options. So if you have:
root=/dev/sda1 -- /foo bar=yes
The kernel will stop at the "--" and pass the rest to userland. But if
you look for "bar", util-linux will incorrectly return "yes".
Ultimately, there's no way for util-linux to exactly parse the command
line the same way as the kernel -- we don't know exactly which ones the
kernel picks up and which it passes on to userland (either as env vars
or as command line args). The kernel passes all unrecognized options.
These updates are simple best effort.
URL: https://bugs.gentoo.org/526754 Signed-off-by: Mike Frysinger <vapier@gentoo.org>
Sami Kerola [Sun, 26 Oct 2014 22:07:08 +0000 (22:07 +0000)]
bash-completion: stop being clever when listing block devices
Various commands such as blkid, cfdisk, fdisk, delpart, and so on listed
only partitions and missed for example disks and volume groups. The
right thing to do is to list all block devices in all for all commands
performing operations with them. This might occasionally list unexpected
devices that I think is lesser bad than missing some.
Addresses: https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=764488 Signed-off-by: Sami Kerola <kerolasa@iki.fi>
Sami Kerola [Sun, 26 Oct 2014 22:07:07 +0000 (22:07 +0000)]
fstrim: initialize path variable
Reading the code this initialization is not necessary, so the change is
mostly about silencing a compiler warning and ensuring the program keeps
on working if there is a future bug that could make uninitialized 'path'
reachable to use.
sys-utils/fstrim.c:247: warning: 'path' may be used uninitialized in this
function
Sami Kerola [Sun, 26 Oct 2014 22:07:06 +0000 (22:07 +0000)]
sfdisk: return deterministic value from sfdisk_deinit()
disk-utils/sfdisk.c:222:9: warning: variable 'rc' is uninitialized when
used here [-Wuninitialized]
return rc;
disk-utils/sfdisk.c:208:8: note: initialize the variable 'rc' to silence
this warning
int rc;
Mike Frysinger [Fri, 24 Oct 2014 16:26:12 +0000 (12:26 -0400)]
bash-completion: fix runuser install
The runuser symlink used to depend on su being enabled, but a refactoring
broke that. So if you build with runuser enabled but not su, you end up
with a broken symlink. Rework the logic so it works in both cases.
To facilitate the calculation of 'cold' vs 'warm' Hardware Clock drift
factor the limit on the update period needs to be less than 8 hours.
4 hours should be enough drift to allow calculations that are not
grossly out of range.
For example, with a workstation that is shutdown every night the cold
drift factor can be significantly different than a drift factor based on
a 24 hour period.
Signed-off-by: J William Piggott <elseifthen@gmx.com>
projects / thirdparty / util-linux.git / log
