Due to either bugs in send (kernel) that generate a command against
a wrong path for example, or transient errors on the receiving side,
we stopped processing the send stream immediately and exited with
an error.
It's often desirable to continue processing the send stream even if an
error happens while processing a single command from the send stream.
This change just adds a --max-errors <N> parameter, whose default value
is 1 (preserving current behaviour), that allows to tolerate N errors
before stopping. A value of 0 means to never stop no matter how many
errors we get into while processing the send stream. Regardless of its
value, errors are always printed to stderr when they happen, just like
before this change.
Signed-off-by: Filipe David Borba Manana <fdmanana@gmail.com>
Signed-off-by: David Sterba <dsterba@suse.cz>
A mdrestore_struct was being written to without its mutex being held.
This race was found with ThreadSanitizer; the relevant part of the report
looks like this:
WARNING: ThreadSanitizer: data race (pid=18828)
Write of size 8 at 0x7fffffc3d088 by main thread:
#0 build_chunk_tree .../btrfs-progs/btrfs-image.c:2233
#1 __restore_metadump .../btrfs-progs/btrfs-image.c:2294
#2 restore_metadump .../btrfs-progs/btrfs-image.c:2345
#3 main .../btrfs-progs/btrfs-image.c:2545
Previous read of size 8 at 0x7fffffc3d088 by thread T1 (mutexes: write M0):
#0 restore_worker .../btrfs-progs/btrfs-image.c:1636
Location is stack of main thread.
Mutex M0 created at:
#0 pthread_mutex_init ??:0
#1 mdrestore_init .../btrfs-progs/btrfs-image.c:1766
#2 __restore_metadump .../btrfs-progs/btrfs-image.c:2286
#3 restore_metadump .../btrfs-progs/btrfs-image.c:2345
#4 main .../btrfs-progs/btrfs-image.c:2545
Thread T1 (tid=18830, running) created by main thread at:
#0 pthread_create ??:0
#1 mdrestore_init .../btrfs-progs/btrfs-image.c:1784
#2 __restore_metadump .../btrfs-progs/btrfs-image.c:2286
#3 restore_metadump .../btrfs-progs/btrfs-image.c:2345
#4 main .../btrfs-progs/btrfs-image.c:2545
Signed-off-by: Adam Buchbinder <abuchbinder@google.com>
Signed-off-by: David Sterba <dsterba@suse.cz>
Allow the specification of the filesystem UUID at mkfs time.
Non-unique unique IDs are rejected. This includes attempting
to re-mkfs with the same UUID; if you really want to do that,
you can mkfs with a new UUID, then re-mkfs with the one you
wanted.
(Implemented only for mkfs.btrfs, not btrfs-convert).
Signed-off-by: Eric Sandeen <sandeen@redhat.com>
[converted help to asciidoc]
Signed-off-by: David Sterba <dsterba@suse.cz>
gcc 4.9.0 gives a warning: format ‘%d’ expects argument of type ‘int’,
but argument 2 has type ‘u64’
Using %llu and casting to unsigned long long (same as bytenr) fixes this.
Signed-off-by: Christian Hesse <mail@eworm.de>
Signed-off-by: David Sterba <dsterba@suse.cz>
When I made all the btrfs-foo.c targets generic, I somehow
managed to break the libs definition for btrfs-fragments by
dropping the "s" off the end.
Fix that, although apparently nobody is building this tool. :)
Signed-off-by: Eric Sandeen <sandeen@redhat.com>
Signed-off-by: David Sterba <dsterba@suse.cz>
The original print_dev_item() only prints device id,total bytes and
bytes used.
When it comes to debug duplicated device id, dev uuid
is needed to distinguish different devices since device id is not
reliable.
(Although the current dev replace implement will reuse the dev uuid,
so not really helpful)
This patch adds dev uuid output for print_dev_item().
Signed-off-by: Qu Wenruo <quwenruo@cn.fujitsu.com>
Signed-off-by: David Sterba <dsterba@suse.cz>
The btrfstune -S option accepts a positive value to enable seeding,
and a zero to disable seeding, negtive is not allowed.
Add "positive, zero, negative" sentences to btrfstune manpage.
Signed-off-by: Gui Hecheng <guihc.fnst@cn.fujitsu.com>
Signed-off-by: David Sterba <dsterba@suse.cz>
Fedora had a bug where a poor user thought that --alloc-start
meant that the filesystem would be created at that offset into
the device, rather than just starting allocations at that offset.
A subtle difference, but worth clarifying, because the manpage
is misleading on this point.
The original commit log for this option says:
Add mkfs.btrfs -A offset to control allocation start on devices
This is a utility option for the resizer, it makes sure to allocate
at offset bytes in the disk or higher. It ensures the resizer will have
something to move when testing it.
so allude to that intended use in the manpage.
Signed-off-by: Eric Sandeen <sandeen@redhat.com>
[converted to asciidoc]
Signed-off-by: David Sterba <dsterba@suse.cz>
The manpage of btrfsck.8 is supposed to link to btrfs-check.8 .
Reported-by: WorMzy Tykashi <wormzy.tykashi@gmail.com>
Signed-off-by: David Sterba <dsterba@suse.cz>
man pages for btrfs-progs are compressed by gzip by default. In Makefile
the variable GZIP is use, this evaluates to 'gzip gzip' on my system.
>From man gzip:
> The environment variable GZIP can hold a set of default options for
> gzip. These options are interpreted first and can be overwritten by
> explicit command line parameters.
So using any other variable name fixes this.
Signed-off-by: Christian Hesse <mail@eworm.de>
Signed-off-by: David Sterba <dsterba@suse.cz>
Very often while debugging filesystems with many subvolumes and/or
snapshots, specially when they are large, I want to see only the
content of one of the trees. So this change just adds an option
to btrfs-debug-tree to allow to specify the id of the tree we're
interesting in dumping to stdout.
Example: btrfs-debug-tree -t 257 /dev/sdc
Will only dump the tree of the first snapshot or subvolume that was
created.
Signed-off-by: Filipe David Borba Manana <fdmanana@gmail.com>
Signed-off-by: David Sterba <dsterba@suse.cz>
If we point btrfs-show-super at a not-btrfs-device and
try to print all superblocks, bad things are apt to happen:
superblock: bytenr=274877906944, device=/dev/sdc2
---------------------------------------------------------
btrfs-show-super: ctree.h:1984: btrfs_super_csum_size: \
Assertion `!(t >= (sizeof(btrfs_csum_sizes) / sizeof((btrfs_csum_sizes)[0])))' failed.
csum 0xAborted
Don't try to print superblocks that don't look like superblocks,
and add an "-f" (force) option to try anyway, if the user
really wants to give it a shot.
Fix some spelling & capitalization while we're at it.
The manpage says that if any problem happens, 1 will
be returned, but that's already not true today LOL, so
I didn't bother to make it true when we detect bad
sb magic, either...
I figure it's worth continuing and trying all superblocks
in case just one has a corrupt magic.
Signed-off-by: Eric Sandeen <sandeen@redhat.com>
[renamed -f to -F due to clash with existing option, converted
relevant docs to asciidoc]
Signed-off-by: David Sterba <dsterba@suse.cz>
The btrfs-rescue accepts exactly one arg for both
chunk-recover & super-recover, use check_argc_exact clearly.
Signed-off-by: David Sterba <dsterba@suse.cz>
Add sys chunk array and backup roots info if the new option '-f'
if specified.
This may be useful for debugging sys_chunk related issues.
Signed-off-by: Gui Hecheng <guihc.fnst@cn.fujitsu.com>
Signed-off-by: David Sterba <dsterba@suse.cz>
find_good_parent() uses assert to deal with the problem that clone
source's parent can't be found.
But in fact the assert is somewhat overkilled since subvol_uuid_search()
has enough error messages for debug and caller of find_good_parent() can
handle the problems in find_good_parent(), so the assert can be removed
without any problem.
Signed-off-by: Qu Wenruo <quwenruo@cn.fujitsu.com>
Signed-off-by: David Sterba <dsterba@suse.cz>
When 'btrfs replace status' encounters an unknown dev replace status, it
will cause an assert, which is somewhat overkilled and can be replaced
with a normal error message.
Signed-off-by: Qu Wenruo <quwenruo@cn.fujitsu.com>
Signed-off-by: David Sterba <dsterba@suse.cz>
open_path_or_dev_mnt() is used to on *mounted* btrfs device or mount
point, when a unmounted btrfs device is passed, errno is set to EINVAL to
info the caller.
If ignore the errno and just print "ERROR: can't access '%s'", end users
will get confused.
This patch will add check for open_path_or_dev_mnt() caller and print
more meaningful error message when a unmounted btrfs device path is
given.
Signed-off-by: Qu Wenruo <quwenruo@cn.fujitsu.com>
Signed-off-by: David Sterba <dsterba@suse.cz>
The asciidoc conversion was done on a development branch and there are
portions of text that do not reflect the code.
Signed-off-by: David Sterba <dsterba@suse.cz>
The mkfs.btrfs --features long option takes an argument but does not
declare it. Consequently getopt does not allocate an argument, which
makes an unconditional strdup() crash during options parsing.
Fix by declaring the argument in the options alias array.
Signed-off-by: Holger Hoffstätte <holger.hoffstaette@googlemail.com>
Signed-off-by: David Sterba <dsterba@suse.cz>
Remove ununsed parameters since 71d6bd3c in create_raid_groups.
Signed-off-by: Rakesh Pandit <rakesh@tuxera.com>
Signed-off-by: David Sterba <dsterba@suse.cz>
Set string "xattr_name" 's end with '\0' so that it won't be
violated in memory.
With this fix, xfstest/btrfs/048 can pass on my box.
Signed-off-by: Liu Bo <bo.li.liu@oracle.com>
Reviewed-by: Filipe Manana <fdmanana@gmail.com>
Signed-off-by: David Sterba <dsterba@suse.cz>
Originally only if 'block_only' is specified, the 'fs_root == NULL'
will be checked. But if 'block_only' is not specified and close_root
will be called blindly without checking 'fs_root == NULL', which is
unsafe.
Signed-off-by: Gui Hecheng <guihc.fnst@cn.fujitsu.com>
Signed-off-by: David Sterba <dsterba@suse.cz>
Linking with libbtrfs fails because arg_strtou64 is not defined and we
cannot just add utils.o to library objects because it's not
library-clean.
Reported-by: Arvin Schnell <aschnell@suse.com>
Reported-by: Anton Farygin <rider@altlinux.org>
Signed-off-by: David Sterba <dsterba@suse.cz>
For RAID0,5,6,10,
For system chunk, there shouldn't be too many stripes to
make a btrfs_chunk that exceeds BTRFS_SYSTEM_CHUNK_ARRAY_SIZE
For data/meta chunk, there shouldn't be too many stripes to
make a btrfs_chunk that exceeds a leaf.
Signed-off-by: Gui Hecheng <guihc.fnst@cn.fujitsu.com>
Signed-off-by: David Sterba <dsterba@suse.cz>
For system chunk array,
We copy a "disk_key" and an chunk item each time,
so there should be enough space to hold both of them,
not only the chunk item.
Signed-off-by: Gui Hecheng <guihc.fnst@cn.fujitsu.com>
Signed-off-by: David Sterba <dsterba@suse.cz>
In utils.c, zero_end is used as a parameter, should not force it to 1.
In mkfs.c, zero_end is set to 1 or 0(-b) at the beginning, should not
force it to 1 unconditionally.
Signed-off-by: Li Yang <liyang.fnst@cn.fujitsu.com>
Signed-off-by: David Sterba <dsterba@suse.cz>
Fix double free of memory if btrfs_open_devices fails:
*** Error in `btrfs': double free or corruption (fasttop): 0x000000000066e020 ***
Crash happened because when open failed on device inside
btrfs_open_devices it freed all memory by calling btrfs_close_devices but
inside disk-io.c we call btrfs_close_again it again.
Signed-off-by: Rakesh Pandit <rakesh@tuxera.com>
Signed-off-by: David Sterba <dsterba@suse.cz>
It complains errno never gets assigned to zero in find-root and since
errno anyway is zero at program started up, lets remove it.
Check "copy is less then zero" isn't possible because strtoull used by
arg_strtou64 wouldn't return -ve number.
Trivial space fixes.
Signed-off-by: Rakesh Pandit <rakesh@tuxera.com>
Signed-off-by: David Sterba <dsterba@suse.cz>
"close_inode_for_write" always returns 0, so just remove its return
value and remove dead checking in caller.
Signed-off-by: Rakesh Pandit <rakesh@tuxera.com>
Signed-off-by: David Sterba <dsterba@suse.cz>
Lets use "errors" instead of "error" because more then one ref errors
are possible. Also print error messages for unresolved refs in
check_root_refs.
Signed-off-by: Rakesh Pandit <rakesh@tuxera.com>
Signed-off-by: David Sterba <dsterba@suse.cz>
Because the function open_file_or_dir() always opened the input file in
read/write mode (O_RDWR), we were not able to due a compression property
get against a file living in a read-only subvolume/snapshot.
Fix this by opening the file with O_RDONLY mode if we're doing a property
get.
Signed-off-by: Filipe David Borba Manana <fdmanana@gmail.com>
Reviewed-by: Liu Bo <bo.li.liu@oracle.com>
Signed-off-by: David Sterba <dsterba@suse.cz>
It was added in 25d82d22 but broke recently in 4724d7b0 while making
discard interruptible.
Signed-off-by: Rakesh Pandit <rakesh@tuxera.com>
Signed-off-by: David Sterba <dsterba@suse.cz>
When executing 'btrfs prop get' on a file which is not compressed, return
value will always be 50 since prop_compress() return -ENOATTR.
But the codes have already check the errno to avoid unnecessary error
message, so the return value should also set to 0.
Signed-off-by: Qu Wenruo <quwenro@cn.fujitsu.com>
Reviewed-by: Filipe Manana <fdmanana@gmail.com>
Signed-off-by: David Sterba <dsterba@suse.cz>
btrfs_scan_kernel() is only used in 'btrfs fi show' but returns wrong
return value. When search parameter is passed, it will never return 0
even the search can be matched.
This patch will change the whatever strange logic to a more easy to
understand one using 'found' var.
Reported-by: Tsutomu Itoh <t-itoh@jp.fujitsu.com>
Signed-off-by: Qu Wenruo <quwenruo@cn.fujitsu.com>
Cc: Anand Jain <anand.jain@oracle.com>
Signed-off-by: David Sterba <dsterba@suse.cz>
Free memory if open call fails. Prevent pthread_cancel on threads
which have already finished successfully. If all calls to
pthread_create and pthread_join are successful, we mistakenly call
pthread_cancel because cancel_from and cancel_to are both zero.
Make POSIX.1-2001 happy by supplying a non-NULL second argument to
pthread_setcanceltype.
Signed-off-by: Rakesh Pandit <rakesh@tuxera.com>
Signed-off-by: David Sterba <dsterba@suse.cz>
scrub_progress_cycle thread runs in asynchronous type but locks mutex
while reading shared data. This patch disables cancelability for a
brief time while locks are on so as to make sure they are unlocked
before thread is canceled.
scrub_write_progress gets called from scrub_progress_cycle in
asynchronous thread but cancelability is disabled after mutex is
locked. This patch moves the call to set cancelability type before
mutex lock and makes corresponding changes to labels for error
handling.
Signed-off-by: Rakesh Pandit <rakesh@tuxera.com>
Signed-off-by: David Sterba <dsterba@suse.cz>
Threads always use default attributes in all tools, so pthread
attribute objects and their initializations are of no use. Just pass
NULL as attr attribute to pthread_create for default attributes.
Signed-off-by: Rakesh Pandit <rakesh@tuxera.com>
Signed-off-by: David Sterba <dsterba@suse.cz>
If pthread_create fails in mdrestore_init, then number of threads
created could be less then num of threads option. Hence pass number of
successful pthread_create calls to mdrestore_destroy, so that we don't
call pthread_join on IDs not present when pthread_create fails.
metadump_init already had this fixed, but repeats code from
metadump_destroy. Reuse metadump_destroy by passing number of threads
created (successful pthread_create calls) and save repeated cleaup
code. Had to move metadump_destroy before metadump_init for obvious
reasons.
Signed-off-by: Rakesh Pandit <rakesh@tuxera.com>
Signed-off-by: David Sterba <dsterba@suse.cz>
The 'btrfsck' command has been deprecated in favor of 'btrfs check'. For
compatibility install a symlink to the btrfs-check.8 manpage.
Signed-off-by: David Sterba <dsterba@suse.cz>
Regenerating the asciidoc takes much longer now and makes quick build
tests long. There's separate clean-doc target for that and clean-all
that cleans docs and sources.
Signed-off-by: David Sterba <dsterba@suse.cz>
Add btrfs wiki page reference to btrfs-check/btrfsck, btrfs-restore and
btrfs-device as supplement.
Signed-off-by: Qu Wenruo <quwenruo@cn.fujitsu.com>
Signed-off-by: David Sterba <dsterba@suse.cz>
Mark Fasheh