test_dev_for_mkfs() is a common place where
we check if a device is fit for the btrfs use.
cmd_start_replace() should make use of test_dev_for_mkfs(),
and here the test_dev_for_mkfs() is further enhanced
to fit the cmd_start_replace() needs.
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>
Port of commit b3b4aa7 to userspace.
parameter tree root it's not used since commit
5f39d397dfbe140a14edecd4e73c34ce23c4f9ee ("Btrfs: Create extent_buffer
interface for large blocksizes")
This gets userspace a tad closer to kernelspace by removing
this unused parameter that was all over the codebase...
Signed-off-by: Eric Sandeen <sandeen@redhat.com>
Signed-off-by: David Sterba <dsterba@suse.cz>
Signed-off-by: Chris Mason <chris.mason@fusionio.com>
when we scan /proc/partitions the cdrom is scanned
as well, and we don't have to report ENOMEDIUM errors
against it.
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>
This would help to reuse the function
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 dev scan to find btrfs is performed at two locations
all most the same way one at filesystem show and another
at device scan. They both follow the same steps. This
patch does not alter anything except that it brings these
two same logic into the function scan_for_btrfs so that
we can play tweaking it.
the patch which recommends to use /dev/mapper
will also need it
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>
btrfs_scan_for_fsid uses only one argument run_ioctl out of 3
so remove the rest two of them
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>
valgrind complains open_file_or_dir() causes a memory leak.That is because
if we open a directoy by opendir(), and then we should call closedir()
to free memory.
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>
As implemented now, we use 1024 based units but reporting 1000 based,
let's finally fix that and add optional unit bases later.
Signed-off-by: David Sterba <dsterba@suse.cz>
Signed-off-by: Chris Mason <chris.mason@fusionio.com>
Instead of aborting with a BUG_ON() statement, return a
negated errno code. Also updated mkfs and convert tools
to print a nicer error message when make_btrfs() returns
an error.
Signed-off-by: Filipe David Borba Manana <fdmanana@gmail.com>
Signed-off-by: David Sterba <dsterba@suse.cz>
Signed-off-by: Chris Mason <chris.mason@fusionio.com>
Assert that the writes of the device and chunk tree
roots succeed. This verification is currently done
for all other tree roots, however it was missing for
those 2 trees.
Would these tree root writes fail, but all others succeed,
it would lead to a corrupted/incomplete btrfs filesystem,
or, more likely some weird failure later on in mkfs.btrfs
inside open_ctree().
Signed-off-by: Filipe David Borba Manana <fdmanana@gmail.com>
Signed-off-by: David Sterba <dsterba@suse.cz>
Signed-off-by: Chris Mason <chris.mason@fusionio.com>
We don't need callers to manage string storage for each pretty_sizes()
call. We can use a macro to have per-thread and per-call static storage
so that pretty_sizes() can be used as many times as needed in printf()
arguments without requiring a bunch of supporting variables.
This lets us have a natural interface at the cost of requiring __thread
and TLS from gcc and a small amount of static storage. This seems
better than the current code or doing something with illegible format
specifier macros.
Signed-off-by: Zach Brown <zab@redhat.com>
Acked-by: Wang Shilong <wangs.fnst@cn.fujitsu.com>
Signed-off-by: David Sterba <dsterba@suse.cz>
Signed-off-by: Chris Mason <chris.mason@fusionio.com>
Some codes still use the cpu_to_lexx instead of the
BTRFS_SETGET_STACK_FUNCS declared in ctree.h.
Also added some BTRFS_SETGET_STACK_FUNCS for btrfs_header and
btrfs_super.
Signed-off-by: Qu Wenruo <quwenruo@cn.fujitsu.com>
Signed-off-by: David Sterba <dsterba@suse.cz>
Signed-off-by: Chris Mason <chris.mason@fusionio.com>
With commit
87c09f7 Btrfs-progs: fix memory leaks on cleanup
mkfs on multiple dev is ending with segfault at
close_all_devices() during kfree(device->name)
because mkfs calls btrfs_add_to_fsid, which does not initialize
name when dev is added to the list.
Signed-off-by: Anand Jain <anand.jain@oracle.com>
Signed-off-by: Chris Mason <chris.mason@fusionio.com>
In the cases where one of the disk is not suitable for
btrfs, then we would fail the mkfs, however we determine
that after we have written btrfs to the preceding disks.
At this time if user changes mind for not to use btrfs
will left with no choice.
So this patch will check if all the provided disks are
suitable for the btrfs at once before proceeding to
create btrfs on a disk.
Further this patch also removed duplicate code to check
device suitability for the btrfs.
Next, there is an existing bug about the -r mkfs option,
which this patch would carry forward most of it.
Ref:
[PATCH 2/2, RFC] btrfs-progs: overhaul mkfs.btrfs -r option
Signed-off-by: Anand Jain <anand.jain@oracle.com>
to merg prev
Signed-off-by: Anand Jain <anand.jain@oracle.com>
Previously btrfs-image would set a METADUMP flag and would make one big system
chunk to cover the entire file system in the super in order to get around the
unpleasant business of having to adjust the chunk tree. This meant that you
could use the progs stuff on a restored file system, which is great for testing
btrfsck and other such things. But we want to be able to run the tree log
replay on a file system that is not able to run the tree log replay. So in
order to do this we need to fixup the super's chunk array and the chunk tree
itself. This is pretty easy since we restore using the logical offsets of the
metadata, so we just have to set the chunk items to have 1 stripe and have the
stripes point at the primary device and then use the logical offset of the chunk
as the physical offset. With this patch I can restore a file system image that
had a tree log and mount the file system and have the log be replayed
successfully. This patch also gives you the -o option in case you want the old
restore way, in the case where we want to make sure the system chunks as they
were given to us are correct. Thanks,
Signed-off-by: Josef Bacik <jbacik@fusionio.com>
Of recently and intermittently I am seeing open fail
for /dev/btrfs-control (btrfs is loaded), and there are no
dmesg errors, this may not be a complete help in digging
this issue but something which is necessary.
Thanks
Signed-off-by: Anand Jain <anand.jain@oracle.com>
Signed-off-by: David Sterba <dsterba@suse.cz>
get_fs_info() has been silently switching from a device to a mounted
path as needed; the caller's filehandle was unexpectedly closed &
reopened outside the caller's scope. Not so great.
The callers do want "fdmnt" to be the filehandle for the mount point
in all cases, though - the various ioctls act on this (not on an fd
for the device). But switching it in the local scope of get_fs_info
is incorrect; it just so happens that *usually* the fd number is
unchanged.
So - use the new helpers to detect when an argument is a block
device, and open the the mounted path more obviously / explicitly
for ioctl use, storing the filehandle in fdmnt.
Then, in get_fs_info, ignore the fd completely, and use the path on
the argument to determine if the caller wanted to act on just that
device, or on all devices for the filesystem.
Affects those commands which are documented to accept either
a block device or a path:
* btrfs device stats
* btrfs replace start
* btrfs scrub start
* btrfs scrub status
Signed-off-by: Eric Sandeen <sandeen@redhat.com>
Add 3 new helpers:
* is_block_device(), to test if a path is a block device.
* get_btrfs_mount(), to get the mountpoint of a device,
if mounted.
* open_path_or_dev_mnt(path), to open either the pathname
or, if it's a mounted btrfs dev, the mountpoint. Useful
for some commands which can take either type of arg.
Signed-off-by: Eric Sandeen <sandeen@redhat.com>
Allocate fs_info::super_copy dynamically of full BTRFS_SUPER_INFO_SIZE
and use it directly for saving superblock to disk.
This fixes incorrect superblock checksum after mkfs.
Signed-off-by: David Sterba <dsterba@suse.cz>
Clean btrfslabel.[c|h] out of the source tree and move those related
functions to utils.[c|h].
CC: Gene Czarcinski <gene@czarc.net>
Signed-off-by: Jie Liu <jeff.liu@oracle.com>
Signed-off-by: David Sterba <dsterba@suse.cz>
Refactor check_label().
- Make it be static at first, this is a preparation step since we'll remove
btrfslabel.[c|h] and move those functions from there to utils.[c|h], we can
do pre-checking against the input label string with it.
- Fix the label length check up from BTRFS_LABEL_SIZE to BTRFS_LABEL_SIZE - 1.
- Kill the check of label contains an invalid character, see below commits for detail:
79e0e445fc
btrfs-progs: kill check for /'s in labels.
Signed-off-by: Jie Liu <jeff.liu@oracle.com>
CC: David Sterba <dsterba@suse.cz>
CC: Gene Czarcinski <gene@czarc.net>
Currently, the following commands succeed.
# cat /proc/swaps
Filename Type Size Used Priority
/dev/sda3 partition 8388604 0 -1
/dev/sdc8 partition 9765884 0 -2
# mkfs.btrfs /dev/sdc8
WARNING! - Btrfs v0.20-rc1-165-g82ac345 IS EXPERIMENTAL
WARNING! - see http://btrfs.wiki.kernel.org before using
fs created label (null) on /dev/sdc8
nodesize 4096 leafsize 4096 sectorsize 4096 size 9.31GB
Btrfs v0.20-rc1-165-g82ac345
# btrfs fi sh /dev/sdc8
Label: none uuid: fc0bdbd0-7eed-460f-b4e9-131273b66df2
Total devices 1 FS bytes used 28.00KB
devid 1 size 9.31GB used 989.62MB path /dev/sdc8
Btrfs v0.20-rc1-165-g82ac345
#
But we should check out the swap device. Fixed it.
Signed-off-by: Tsutomu Itoh <t-itoh@jp.fujitsu.com>
Tested-by: David Sterba <dsterba@suse.cz>
print more informative error when we fail to open a device
If open() fails, we should let the user know why it failed.
Signed-off-by: Eric Sandeen <sandeen@redhat.com>
Signed-off-by: Gene Czarcinski <gene@czarc.net>
In the places where we copy a string into the name
member of btrfs_ioctl_vol_args or btrfs_ioctl_vol_args_v2,
we use strncopy (to not overflow the name array) and then
set the last position to the null character.
Howver, in both cases the arrays are defined with:
char name[MAX+1];
hence the last array position is name[MAX].
In most cases, we now insert the null at name[MAX-1]
which deprives us of one useful character.
Even the above isn't consistent through the code, so
make some helper code to make it simple, i.e.
strncpy_null(dest, src) which automatically does the
right thing based on the size of dest.
Thanks to Zach Brown for the macro suggestion.
Signed-off-by: Eric Sandeen <sandeen@redhat.com>
btrfs_scan_one_dir() can overflow an arbitrarily small 256 byte buffer
with an arbitrarily slightly larger 1024 byte buffer as it remembers the
path of a dir to later descend.
Make these buffers the same size to stop the overflow and chose PATH_MAX
for that size so that it won't fail on legitimately bonkers paths.
Signed-off-by: Zach Brown <zab@redhat.com>
The super block magic is a le64 whose value looks like an unterminated
string in memory. The lack of null termination leads to clumsy use of
string functions and causes static analysis tools to warn that the
string will be unterminated.
So let's just treat it as the le64 that it is. Endian wrappers are used
on the constant so that they're compiled into run-time constants.
Signed-off-by: Zach Brown <zab@redhat.com>
Two convenient utility functions that have so far been local to scrub are
moved to utils.c.
They will be used in the device stats code in a following commit.
Signed-off-by: Stefan Behrens <sbehrens@giantdisaster.de>
The definition of the function open_file_or_dir() is moved from common.c
to utils.c in order to be able to share some common code between scrub
and the device stats in the following step. That common code uses
open_file_or_dir(). Since open_file_or_dir() makes use of the function
dirfd(3), the required XOPEN version was raised from 6 to 7.
Signed-off-by: Anand Jain <anand.jain@oracle.com>
Original-Signed-off-by: Stefan Behrens <sbehrens@giantdisaster.de>
The LOOP_GET_STATUS ioctl truncates filenames to 64 characters. We should get
the backing file for a given loop device from /sys/. This is how losetup does it
as well.
Signed-off-by: Nirbheek Chauhan <nirbheek.chauhan@collabora.co.uk>
Signed-off-by: Gene Czarcinski <gene@czarc.net>
Tested-By: Hector Oron <hector.oron@collabora.co.uk>
Ignore the error ENXIO (device don't exists) and ENOMEDIUM (
No medium found -> like a cd tray empty) in the function
btrfs_scan_one_dir.
This avoids spurios errors due to an empty CD or a block device node
without a device (which is frequent in a static /dev).
Signed-off-by: Goffredo Baroncelli <kreijack@inwind.it>
Add new suffixes in parse_size() function. New suffixes are: T as
terabyte, P as petabyte, E as exabyte. Note these units are
multiply of 2 .
Signed-off-by: Goffredo Baroncelli <kreijack@inwind.it>
Replace the function atoll with strtoull(); Check that the suffix for the
parse_size() input is of only one character.
Signed-off-by: Goffredo Baroncelli <kreijack@inwind.it>
My patch
04609add88
introduced a regression where if you mkfs'ed a group of disks with different
sizes it limited the disks to the size of the first one that is specified.
This was not the intent of my patch, I only want it to limit the size based
on the -b option, so I've reworked the code to pass in a max block count and
that fixes the issue. Thanks,
Signed-off-by: Josef Bacik <jbacik@fusionio.com>
I had a test that creates a 7gig raid1 device but it was ending up wonky
because the second device that gets added is the full size of the disk
instead of the limited size. So enforce the limited size on all disks
passed in at mkfs time, otherwise our threshold calculations end up wonky
when doing chunk allocations. Thanks,
Signed-off-by: Josef Bacik <josef@redhat.com>
If we iterate the "goto again" loop, we've called "closedir(dirp)",
yet at the top of the loop, upon malloc failure we "goto fail",
where we test dirp and if non-NULL, call closedir(dirp) again.
* utils.c (btrfs_scan_one_dir): Clear "dirp" after closedir to avoid
use-after-free upon failed fullpath = malloc(...
Signed-off-by: Jim Meyering <meyering@redhat.com>
When we're using multipath or raid0, it is possible
that btrfs dev scan will find one of the component devices
instead of the proper virtual device the kernel creates.
We want to make sure the kernel scans the virtual devices last,
since it always remembers the last device it finds with a given fsid.
Signed-off-by: Chris Mason <chris.mason@oracle.com>
btrfs_scan_for_fsid is used by open_ctree and by mkfs when it is
checking for mounted devices. It currently scans all of /dev,
which is rarely the right answer.
Signed-off-by: Chris Mason <chris.mason@oracle.com>
/proc/mounts contains device names that don't exist,
we end up erroring out because we're not able to stat
the device (that doesn't exist).
Fix this by allowing the mkfs when the target device doesn't exist.
Signed-off-by: Chris Mason <chris.mason@oracle.com>
