I hit a segfault when deleting a subvolume with very long name(>4096),
it's because cmd_subvol_delete() calls strdup() and passes NULL as
argument, which is returned by realpath(3).
I used the following script to reproduce
#!/bin/bash
mnt=$1
i=1
path=$mnt/subvol_$i
# Create very deep subvolumes
while btrfs sub create $path;do
((i++))
path="$path/subvol_$i"
done
last_vol=$(dirname $path)
dir=$(dirname $last_vol)
vol=$(basename $last_vol)
# Try to delete tha last one, this would get segfault
pushd $dir
btrfs sub delete $vol
popd
Fix it by checking return value of realpath(3), also fix the one in
find_mount_root().
Signed-off-by: Eryu Guan <guaneryu@gmail.com>
Signed-off-by: David Sterba <dsterba@suse.cz>
Signed-off-by: Chris Mason <chris.mason@fusionio.com>
Check for fopen() failure. This shows up in static analysis as a
possible null pointer derference.
Signed-off-by: Zach Brown <zab@redhat.com>
Laughed-at-by: Stefan Behrens <sbehrens@giantdisaster.de>
Signed-off-by: David Sterba <dsterba@suse.cz>
Signed-off-by: Chris Mason <chris.mason@fusionio.com>
This was found by static analysis.
Signed-off-by: Zach Brown <zab@redhat.com>
Signed-off-by: David Sterba <dsterba@suse.cz>
Signed-off-by: Chris Mason <chris.mason@fusionio.com>
Presumably people missed these warnings because btrfs-fragments isn't
built by default.
Signed-off-by: Zach Brown <zab@redhat.com>
Signed-off-by: David Sterba <dsterba@suse.cz>
Signed-off-by: Chris Mason <chris.mason@fusionio.com>
Stop iteration at the number of elements in the colors[] array when
initializing the elements. Rather than a magic number. This was found
by static analysis.
Signed-off-by: Zach Brown <zab@redhat.com>
Signed-off-by: David Sterba <dsterba@suse.cz>
Signed-off-by: Chris Mason <chris.mason@fusionio.com>
commit 4782e8ebdb accidentally replaced
[0] with [-1]. Put it back. This was found by static analysis.
Signed-off-by: Zach Brown <zab@redhat.com>
Reviewed-by: Eric Sandeen <sandeen@redhat.com>
Signed-off-by: David Sterba <dsterba@suse.cz>
Signed-off-by: Chris Mason <chris.mason@fusionio.com>
This was found by static analysis.
Signed-off-by: Zach Brown <zab@redhat.com>
Reviewed-by: chandan <chandan@linux.vnet.ibm.com
Signed-off-by: David Sterba <dsterba@suse.cz>
Signed-off-by: Chris Mason <chris.mason@fusionio.com>
This was found by static analysis.
Signed-off-by: Zach Brown <zab@redhat.com>
Signed-off-by: David Sterba <dsterba@suse.cz>
Signed-off-by: Chris Mason <chris.mason@fusionio.com>
Don't carry around dead code. If its needed again, it's only a few git
commands away. This was found by static analysis.
Signed-off-by: Zach Brown <zab@redhat.com>
Signed-off-by: David Sterba <dsterba@suse.cz>
Signed-off-by: Chris Mason <chris.mason@fusionio.com>
In principle, link_subvol() can be given an abitrary string as the name
of the saved subvolume. It copies it into a fixed-size stack buffer and
then uses it as dirent names without testing its length.
This limits its length to BTRFS_NAME_LEN. This was found by static
analsys.
Signed-off-by: Zach Brown <zab@redhat.com>
Signed-off-by: David Sterba <dsterba@suse.cz>
Signed-off-by: Chris Mason <chris.mason@fusionio.com>
search_for_chunk_blocks() allocates a fixed-size buffer and then reads
arbitrary u32 sized buffers in to it. Instead let's fail if the item is
bigger than the buffer. This was found by static analysis.
Signed-off-by: Zach Brown <zab@redhat.com>
Signed-off-by: David Sterba <dsterba@suse.cz>
Signed-off-by: Chris Mason <chris.mason@fusionio.com>
find_collision() allocates name_len bytes for its sub array so the index
must be less than name_len. This was found by static analysis.
Signed-off-by: Zach Brown <zab@redhat.com>
Signed-off-by: David Sterba <dsterba@suse.cz>
Signed-off-by: Chris Mason <chris.mason@fusionio.com>
btrfs-corrupt-block added some untested path allocations. These showed
up in static analysis when they pass their path to btrfs_search_slot()
which unconditionally dereferences the path.
Signed-off-by: Zach Brown <zab@redhat.com>
Signed-off-by: David Sterba <dsterba@suse.cz>
Signed-off-by: Chris Mason <chris.mason@fusionio.com>
Sometimes, we need to catch length of snprintf() in pretty_size_snprintf().
Signed-off-by: Wang Shilong <wangsl.fnst@cn.fujitsu.com>
Signed-off-by: David Sterba <dsterba@suse.cz>
Signed-off-by: Chris Mason <chris.mason@fusionio.com>
This patch enhance to print the result as a table.
You can use it like:
btrfs qgroup show <path>
However, to table the result better, we make '-p' and '-c' not present
at the same time.
For example:
btrfs qgroup show -pr <path>
The result will output as the follow format:
qgroupid rfer excl max_excl parent
-------- ---- ---- -------- ------
0/265 1289752576 1289752576 0 ---
1/0 0 0 10999511627776 2/0,3/0
2/0 0 0 0 ---
3/0 0 0 0 ---
Signed-off-by: Wang Shilong <wangsl-fnst@cn.fujitsu.com>
Signed-off-by: Miao Xie <miaox@cn.fujitsu.com>
Signed-off-by: David Sterba <dsterba@suse.cz>
Signed-off-by: Chris Mason <chris.mason@fusionio.com>
You might want to list qgroups in order of some items, such as 'qgroupid', 'rfer'
and so on, you can use '--sort'. Now you can sort the qgroups by 'qgroupid',
'rfer','excl','max_rfer' and 'max_excl'.
For example:
If you want to list qgroups in order of 'qgroupid'.
You can use the option like that:
btrfs qgroup show --sort=+/-qgroupid <path>
Here, '+' means the result is sorted by ascending order. '-' is by descending
order. If you don't specify either '+' nor '-', the result is sorted by
default - ascending order.
If you want to combine sort items, you do it like that:
btrfs qgroup show --sort=-qgroupid,+rfer,max_rfer,excl <path>
Signed-off-by: Wang Shilong <wangsl-fnst@cn.fujitsu.com>
Signed-off-by: Miao Xie <miaox@cn.fujitsu.com>
Signed-off-by: David Sterba <dsterba@suse.cz>
Signed-off-by: Chris Mason <chris.mason@fusionio.com>
This patch introduces '-f' option which can help you filter the qgroups
by the path name, you may use it like:
btrfs qgroup show -f <path>
For example:
qgroupid(2/0)
/ \
/ \
qgroupid(1/0)
/ \
/ \
/ \
qgroupid(0/1) qgroupid(0/2)
sub1 sub2
/ \
/ \
dir1 file1
If we use the command:
btrfs qgroup show -f sub1/dir1
The result will output
0/1 -- --
'-f' option helps you list all qgroups impact given path.
(exclude ancestral qgroups)
Signed-off-by: Wang Shilong <wangsl-fnst@cn.fujitsu.com>
Signed-off-by: Miao Xie <miaox@cn.fujitsu.com>
Signed-off-by: David Sterba <dsterba@suse.cz>
Signed-off-by: Chris Mason <chris.mason@fusionio.com>
This patch introduces '-F' option which can help you filter the qgroups
by the path name, you may use it like:
btrfs qgroup show -F <path>
For example:
qgroupid(2/0)
/ \
/ \
qgroupid(1/0)
/ \
/ \
/ \
qgroupid(0/1) qgroupid(0/2)
sub1 sub2
/ \
/ \
dir1 file1
If we use the command:
btrfs qgroup show -F sub1/dir1
The result will output
0/1 -- --
1/0 -- --
2/0 -- --
'-F' option help you list all qgroups impact given path.
(include ancestral qgroups).
Signed-off-by: Wang Shilong <wangsl-fnst@cn.fujitsu.com>
Signed-off-by: Miao Xie <miaox@cn.fujitsu.com>
Signed-off-by: David Sterba <dsterba@suse.cz>
Signed-off-by: Chris Mason <chris.mason@fusionio.com>
This patch introduce '-e' option to print max exclusive size of qgroups.
You may use it like this:
btrfs qgroup -e <path>
Signed-off-by: Wang Shilong <wangsl-fnst@cn.fujitsu.com>
Signed-off-by: Miao Xie <miaox@cn.fujitsu.com>
Signed-off-by: David Sterba <dsterba@suse.cz>
Signed-off-by: Chris Mason <chris.mason@fusionio.com>
This patch introduces '-r' option to print max referenced size of qgroups.
You may use it like:
btrfs qgroup show -r <path>
Signed-off-by: Wang Shilong <wangsl-fnst@cn.fujitsu.com>
Signed-off-by: Miao Xie <miaox@cn.fujitsu.com>
Signed-off-by: David Sterba <dsterba@suse.cz>
Signed-off-by: Chris Mason <chris.mason@fusionio.com>
This patch introduces '-c' option to print the ID of the child qgroups.
You may use it like:
btrfs qgroup show -c <path>
For Example:
qgroupid(2/0)
/ \
/ \
/ \
qgroupid(1/0) qgroupid(1/1)
\ /
\ /
qgroupid(0/1)
If we use the command:
btrfs qgroup show -c <path>
The result will output
0/1 -- -- --
1/0 -- -- 0/1
1/1 -- -- 0/1
2/0 -- -- 1/0,1/1
Signed-off-by: Wang Shilong <wangsl-fnst@cn.fujitsu.com>
Signed-off-by: Miao Xie <miaox@cn.fujitsu.com>
Signed-off-by: David Sterba <dsterba@suse.cz>
Signed-off-by: Chris Mason <chris.mason@fusionio.com>
This patch introduces '-p' option to print the ID of the parent qgroups.
You may use it like:
btrfs qgroup show -p <path>
For Example:
qgroupid(2/0)
/ \
/ \
/ \
qgroupid(1/0) qgroupid(1/1)
\ /
\ /
qgroupid(0/1)
If we use the command:
btrfs qgroup show -p <path>
The result will output
0/1 -- -- 1/0,1/1
1/0 -- -- 2/0
1/1 -- -- 2/0
2/0 -- -- --
Signed-off-by: Wang Shilong <wangsl-fnst@cn.fujitsu.com>
Signed-off-by: Miao Xie <miaox@cn.fujitsu.com>
Signed-off-by: David Sterba <dsterba@suse.cz>
Signed-off-by: Chris Mason <chris.mason@fusionio.com>
The current show_qgroups() just shows a little information, and it is hard to
add some functions which the users need in the future, so i restructure it, make
it easy to add new functions.
In order to improve the scalability of show_qgroups(), i add some important
structures:
struct qgroup_lookup {
struct rb_root root;
}
/*
*store qgroup's information
*/
struct btrfs_qgroup {
struct rb_node rb_node;
u64 qgroupid;
u64 generation;
u64 rfer;
u64 rfer_cmpr;
u64 excl_cmpr;
u64 flags;
u64 max_rfer;
u64 max_excl;
u64 rsv_rfer;
u64 rsv_excl;
struct list_head qgroups;
struct list_head members;
}
/*
*glue structure to represent the relations
*between qgroups
*/
struct btrfs_qgroup_list {
struct list_head next_qgroups;
struct list_head next_member;
struct btrfs_qgroup *qgroup;
struct btrfs_qgroup *member;
}
The above 3 structures are used to manage all the information
of qgroups.
struct {
char *name;
char *column_name;
int need_print;
} btrfs_qgroup_columns[]
We define a arrary to manage all the columns that can be
outputed, and use a member variant(->need_print) to control
the output of the relative column. Some columns are outputed
by default. But we can change it according to the requirement
of the users.
For example:
if outputing max referenced size of qgroup is needed,the function
'btrfs_qgroup_setup_column()' will be called, and the parameter 'BTRFS_QGROUP_MAX_RFER'
(extend in the future) will be passsed to the function. After the function is done,
when showing qgroups, max referenced size of qgroup will be output.
Signed-off-by: Wang Shilong <wangsl-fnst@cn.fujitsu.com>
Signed-off-by: Miao Xie <miaox@cn.fujitsu.com>
Signed-off-by: David Sterba <dsterba@suse.cz>
Signed-off-by: Chris Mason <chris.mason@fusionio.com>
The current code returns from the function when the call to ioctl
fails. This may leak cache_dir_name and cache_full_name. Fix it.
Signed-off-by: chandan <chandan@linux.vnet.ibm.com>
Signed-off-by: David Sterba <dsterba@suse.cz>
Signed-off-by: Chris Mason <chris.mason@fusionio.com>
We usually print out a hex value of any errors on inodes or their backrefs,
which is a huge pain for me because I have to put it into a calculator and count
the bits to figure out which errors these map to, and usually I get it wrong the
first time. To fix this lets just print out a human readable string for each
error that way it will be easier to spot the "OH GOD THAT'S AWFUL" errors from
"oh yeah thats no big deal, repair will fix that." Thanks,
Signed-off-by: Josef Bacik <jbacik@fusionio.com>
Signed-off-by: David Sterba <dsterba@suse.cz>
Signed-off-by: Chris Mason <chris.mason@fusionio.com>
Add the -r option to the manpages for btrfs defragment
Signed-off-by: Frank Holton <fholton@gmail.com>
Signed-off-by: David Sterba <dsterba@suse.cz>
Signed-off-by: Chris Mason <chris.mason@fusionio.com>
A user was reporting an issue with bad transid errors on his blocks. The thing
is that btrfs-progs will ignore transid failures for things like restore and
fsck so we can do a best effort to fix a users file system. So fsck can put
together a coherent view of the file system with stale blocks. So if everything
else is ok in the mind of fsck then we can recow these blocks to fix the
generation and the user can get their file system back. Thanks,
Signed-off-by: Josef Bacik <jbacik@fusionio.com>
Signed-off-by: David Sterba <dsterba@suse.cz>
Signed-off-by: Chris Mason <chris.mason@fusionio.com>
This is a verification test for the transid recow functionality of btrfsck.
I've also adjusted the test script to spit out which image it's testing so I can
be sure the image was getting tested. Thanks,
Signed-off-by: Josef Bacik <jbacik@fusionio.com>
Signed-off-by: David Sterba <dsterba@suse.cz>
Signed-off-by: Chris Mason <chris.mason@fusionio.com>
This patch allows us to garble the generation field of a metadata block. We
will search down to the block, set the bogus generation and write the block back
out. I used this for verifying a transid fix patch for btrfsck. Thanks,
Signed-off-by: Josef Bacik <jbacik@fusionio.com>
Signed-off-by: David Sterba <dsterba@suse.cz>
Signed-off-by: Chris Mason <chris.mason@fusionio.com>
We need to start adding some sanity tests to btrfs-progs to make sure we aren't
breaking things with our patches. The most important of these tools is btrfsck.
This patch gets things started by adding a basic btrfsck test that makes sure we
can fix a corruption problem we know we can fix. Thanks,
Signed-off-by: Josef Bacik <jbacik@fusionio.com>
Signed-off-by: David Sterba <dsterba@suse.cz>
Signed-off-by: Chris Mason <chris.mason@fusionio.com>
Add an option to defrag all files in a directory recursively.
Signed-off-by: Frank Holton <fholton@gmail.com>
Signed-off-by: David Sterba <dsterba@suse.cz>
Signed-off-by: Chris Mason <chris.mason@fusionio.com>
It's just annoying to have to pass it around everywhere. Thanks,
Signed-off-by: Josef Bacik <jbacik@fusionio.com>
Signed-off-by: David Sterba <dsterba@suse.cz>
Signed-off-by: Chris Mason <chris.mason@fusionio.com>
as of now, when 'btrfs device add' adds a device it doesn't
check if the given device contains an existing FS. This
patch will change that to check the same. which when true
add will fail, and ask user to use -f option to overwrite.
further, since now we have test_dev_for_mkfs() function
to check if a disk can be used, so this patch will also
use this function to test the given device before adding.
Signed-off-by: Anand Jain <anand.jain@oracle.com>
Signed-off-by: David Sterba <dsterba@suse.cz>
Signed-off-by: Chris Mason <chris.mason@fusionio.com>
The sync makes sure that 'very recently' introduced delayed work is
accounted for in the output of 'btrfs subvolume find-new' command.
Signed-off-by: chandan <chandan@linux.vnet.ibm.com>
Signed-off-by: David Sterba <dsterba@suse.cz>
Signed-off-by: Chris Mason <chris.mason@fusionio.com>
This is a prepatory work for the btrfs fi show command
fixes. So that we have a function get_df to get the fs sizes
Signed-off-by: Anand Jain <anand.jain@oracle.com>
Signed-off-by: David Sterba <dsterba@suse.cz>
Signed-off-by: Chris Mason <chris.mason@fusionio.com>
Internally, btrfs_header_fsid() calculates an unsigned long, but casts
it to a pointer, while all callers cast it to unsigned long again.
Committed to btrfs as fba6aa75654394fccf2530041e9451414c28084f
Fix line length issues and match changes to kernelspace
Signed-off-by: Ross Kirk <ross.kirk@gmail.com>
Signed-off-by: David Sterba <dsterba@suse.cz>
Signed-off-by: Chris Mason <chris.mason@fusionio.com>
The 'prealloc' extent_state structure is leaked for the case when the 'desired
range' encapsulates/covers the 'extent range'.
Signed-off-by: chandan <chandan@linux.vnet.ibm.com>
Signed-off-by: David Sterba <dsterba@suse.cz>
Signed-off-by: Chris Mason <chris.mason@fusionio.com>
Before I had been dividing by 5 but that gave me too much output so I changed it
to 20 without changing the min seeks test. Fix this to avoid a divide by 0
problem. Thanks,
Signed-off-by: Josef Bacik <jbacik@fusionio.com>
Signed-off-by: David Sterba <dsterba@suse.cz>
Signed-off-by: Chris Mason <chris.mason@fusionio.com>
The message about trim was printed unconditionally, we should check if
trim is supported at all.
Reviewed-by: Eric Sandeen <sandeen@redhat.com>
Signed-off-by: David Sterba <dsterba@suse.cz>
Signed-off-by: Chris Mason <chris.mason@fusionio.com>
I've been wanting to get back to the allocator and make some changes to try and
fix our fragmenation woes with lots of metadata. But in order to make these
changes I need to have something to tell me if my changes are making a real
measurable difference. So this patch adds a bunch of new statistics to
btrfs-calc-size. It will tell me how long it took to read in the trees, how
many seeks it had (both forward and backward). It will tell me how far spread
out the tree is and spit out a nice histogram of the seeks. Here is some sample
output
Calculating size of extent tree
Total size: 60.74MB
Inline data: 0.00
Total seeks: 5020
Forward seeks: 3691
Backward seeks: 1329
Avg seek len: 929.53MB
Seek histogram
4096 - 4096: 1043 ####
8192 - 73728: 760 ###
81920 - 52527104: 753 ###
53518336 - 168009728: 753 ###
168591360 - 696045568: 753 ###
696238080 - 7560364032: 753 ###
7560437760 - 8409739264: 178 |
Total clusters: 1874
Avg cluster size: 25.17KB
Min cluster size: 8.00KB
Max cluster size: 472.00KB
Total disk spread: 7.90GB
Total read time: 0 s 341670 us
Levels: 4
This way we can have good numbers to back up any changes we make to the
allocator. Thanks,
Signed-off-by: Josef Bacik <jbacik@fusionio.com>
Signed-off-by: David Sterba <dsterba@suse.cz>
Signed-off-by: Chris Mason <chris.mason@fusionio.com>
Find the tree id of the containing subvolume for a given file or
directory. For subvolume return it's own id.
$ btrfs inspect-internal rootid <path>
Signed-off-by: David Sterba <dsterba@suse.cz>
Signed-off-by: Chris Mason <chris.mason@fusionio.com>
Remove unused parameter, 'eb'. Unused since introduction in
7777e63b42
Signed-off-by: Ross Kirk <ross.kirk@gmail.com>
Reviewed-by: Eric Sandeen <sandeen@redhat.com>
Signed-off-by: David Sterba <dsterba@suse.cz>
Signed-off-by: Chris Mason <chris.mason@fusionio.com>
Remove unused eb parameter from btrfs_item_nr, unused since introduced
in 7777e63b42
Signed-off-by: Ross Kirk <ross.kirk@gmail.com>
Reviewed-by: Eric Sandeen <sandeen@redhat.com>
Signed-off-by: David Sterba <dsterba@suse.cz>
Signed-off-by: Chris Mason <chris.mason@fusionio.com>
Until now if one of device's first superblock is corrupt,btrfs will
fail to mount. Luckily, btrfs have at least two superblocks for
every disk.
In theory, if silent corrupting happens when we are writting superblocks
into disk, we must hold at least one good superblock.
One side effect is that user must gurantee that the disk must be
a btrfs disk. Otherwise, this tool may destroy other fs.(This is also
reason why btrfs only use first superblock in every disk to mount)
This little program will try to correct bad superblocks from
good superblocks with max generation.
There will be five kinds of return values:
0: all supers are valid, no need to recover
1: usage or syntax error
2: recover all bad superblocks successfully
3: fail to recover bad superblocks
4: abort to recover bad superblocks
Signed-off-by: Wang Shilong <wangsl.fnst@cn.fujitsu.com>
Signed-off-by: David Sterba <dsterba@suse.cz>
Signed-off-by: Chris Mason <chris.mason@fusionio.com>
Signed-off-by: Wang Shilong <wangsl.fnst@cn.fujitsu.com>
Signed-off-by: David Sterba <dsterba@suse.cz>
Signed-off-by: Chris Mason <chris.mason@fusionio.com>
If some fatal superblocks are damaged, running ioctl will return failure,
in this case, we should avoid run ioctl.
Signed-off-by: Wang Shilong <wangsl.fnst@cn.fujitsu.com>
Signed-off-by: David Sterba <dsterba@suse.cz>
Signed-off-by: Chris Mason <chris.mason@fusionio.com>
We don't need to run ioctls when checking whether btrfs
has mounted somewhere.
Signed-off-by: Wang Shilong <wangsl.fnst@cn.fujitsu.com>
Signed-off-by: David Sterba <dsterba@suse.cz>
Signed-off-by: Chris Mason <chris.mason@fusionio.com>
The devices in 'btrfs filesystem show' are now sorted by the device id,
currently the order was undefined.
Signed-off-by: David Sterba <dsterba@suse.cz>
Signed-off-by: Chris Mason <chris.mason@fusionio.com>
Eryu Guan