The new '-b' option will be responsible for converting to block group
tree compat ro feature.
The workflow looks like this for new convert:
- Setting CHANGING_BG_TREE flag
And initialize fs_info->last_converted_bg_bytenr value to (u64)-1.
Any bg with bytenr >= last_converted_bg_bytenr will have its bg item
update go to the new root (bg tree).
- Iterate each block group by their bytenr in descending order
This involves:
* Delete the old bg item from the old tree (extent tree)
* Update last_converted_bg_bytenr to the bytenr of the bg
* Add the new bg item into the new tree (bg tree)
* If we have converted a bunch of bgs, commit current transaction
- Clear CHANGING_BG_TREE flag
And set the new BLOCK_GROUP_TREE compat ro flag and commit.
And since we're doing the convert in multiple transactions, we also need
to resume from last interrupted convert.
In that case, we just grab the last unconverted bg, and start from it.
And to co-operate with the new kernel requirement for both no-holes and
free-space-tree features, the convert tool will check for
free-space-tree feature. If not enabled, will error out with an error
message to how to continue (by mounting with "-o space_cache=v2").
For missing no-holes feature, we just need to set the flag during
convert.
Signed-off-by: Qu Wenruo <wqu@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Similar to kernel check_leaf(), calling btrfs_item_end_nr() may get a
reasonable value even an item has invalid offset/size due to u32
overflow.
Fix it by use u64 variable to store item data end in btrfs_check_leaf()
to avoid u32 overflow.
Bugzilla: https://bugzilla.kernel.org/show_bug.cgi?id=215299
Reported-by: Wenqing Liu <wenqingliu0120@gmail.com>
Signed-off-by: Su Yue <l@damenly.su>
Signed-off-by: David Sterba <dsterba@suse.com>
Those two members are a shortcut for non-RAID56 profiles.
But we should not use such shortcut, and move all our logical address
read/write to the unified read_data_from_disk()/write_data_to_disk().
With previous refactors, now we're safe to remove them.
Signed-off-by: Qu Wenruo <wqu@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
When we change the size of the btrfs_header we're going to need to
change how these helpers calculate where to find the start of items or
block ptrs. To prepare for that make these helpers take the
extent_buffer as an argument so we can do the appropriate math based on
the version type.
Reviewed-by: Nikolay Borisov <nborisov@suse.com>
Signed-off-by: Josef Bacik <josef@toxicpanda.com>
Signed-off-by: David Sterba <dsterba@suse.com>
btrfs_item_nr_offset(0) is simply offsetof(struct btrfs_leaf, items),
which is the same thing as btrfs_leaf_data(), so replace all calls of
btrfs_item_nr_offset(0) with btrfs_leaf_data().
Reviewed-by: Nikolay Borisov <nborisov@suse.com>
Signed-off-by: Josef Bacik <josef@toxicpanda.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Now that all callers are using the _nr variations we can simply rename
these helpers to btrfs_item_##member/btrfs_set_item_##member and change
the actual item SETGET funcs to raw_item_##member/set_raw_item_##member
and then change all callers to drop the _nr part.
Reviewed-by: Nikolay Borisov <nborisov@suse.com>
Signed-off-by: Josef Bacik <josef@toxicpanda.com>
Signed-off-by: David Sterba <dsterba@suse.com>
All callers use the btrfs_item_end_nr() variation, simply drop
btrfs_item_end() and make btrfs_item_end_nr() use the _nr() variations
of the item get helpers.
Reviewed-by: Nikolay Borisov <nborisov@suse.com>
Signed-off-by: Josef Bacik <josef@toxicpanda.com>
Signed-off-by: David Sterba <dsterba@suse.com>
We have a lot of the following patterns
item = btrfs_item_nr(nr);
btrfs_set_item_*(eb, item, val);
btrfs_set_item_*(eb, btrfs_item_nr(nr), val);
in a lot of places in our code. Instead add _nr variations of these
helpers and convert all of the users to this new helper.
Reviewed-by: Nikolay Borisov <nborisov@suse.com>
Signed-off-by: Josef Bacik <josef@toxicpanda.com>
Signed-off-by: David Sterba <dsterba@suse.com>
We have this pattern in a lot of places
item = btrfs_item_nr(slot);
btrfs_item_size(leaf, item);
btrfs_item_offset(leaf, item);
when we could simply use
btrfs_item_size_nr(leaf, slot);
btrfs_item_offset_nr(leaf, slot);
Fix all callers of btrfs_item_size() and btrfs_item_offset() to use the
_nr variation of the helpers.
Reviewed-by: Nikolay Borisov <nborisov@suse.com>
Signed-off-by: Josef Bacik <josef@toxicpanda.com>
Signed-off-by: David Sterba <dsterba@suse.com>
This helper only takes the nodesize, but in the future it'll take a bool
to indicate if we're extent tree v2. The remaining users are all where
we only have extent_buffer, but we should always have a valid
eb->fs_info in these cases, so add BUG_ON()'s for the !eb->fs_info case
and then convert these callers to use BTRFS_LEAF_DATA_SIZE which takes
the fs_info.
Reviewed-by: Nikolay Borisov <nborisov@suse.com>
Signed-off-by: Josef Bacik <josef@toxicpanda.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Unlike kernel, we have cached physical address of extent_buffer in
dev_bytenr. Print it for better debug experience.
Signed-off-by: Su Yue <l@damenly.su>
Signed-off-by: David Sterba <dsterba@suse.com>
Make the helpers using crc32c not inline so the crc32c.h can be removed
from the public headers exported by libbtrfs.
Signed-off-by: David Sterba <dsterba@suse.com>
This function lies in the kernel-shared directory and is supposed to be
close to 1:1 copy with its kernel counterpart, yet it takes one extra
argument - root. But this is now unused to simply remove it.
Reviewed-by: Qu Wenruo <wqu@suse.com>
Signed-off-by: Nikolay Borisov <nborisov@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
The function already takes an extent_buffer which has a reference to
the owning filesystem's fs_info. This also brings the function in line
with the kernel's signature.
Reviewed-by: Qu Wenruo <wqu@suse.com>
Signed-off-by: Nikolay Borisov <nborisov@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
It's not used, so just remove it.
Reviewed-by: Qu Wenruo <wqu@suse.com>
Signed-off-by: Nikolay Borisov <nborisov@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Enabling quota in zoned mored hits the following assertion:
$ mkfs.btrfs -f -d single -m single -R quota /dev/nullb0
btrfs-progs v5.11
See http://btrfs.wiki.kernel.org for more information.
Zoned: /dev/nullb0: host-managed device detected, setting zoned feature
Resetting device zones /dev/nullb0 (1600 zones) ...
bad tree block 25395200, bytenr mismatch, want=25395200, have=0
kernel-shared/disk-io.c:549: write_tree_block: BUG_ON `1` triggered, value 1
./mkfs.btrfs(+0x26aaa)[0x564d1a7ccaaa]
./mkfs.btrfs(write_tree_block+0xb8)[0x564d1a7cee29]
./mkfs.btrfs(__commit_transaction+0x91)[0x564d1a7e3740]
./mkfs.btrfs(btrfs_commit_transaction+0x135)[0x564d1a7e39aa]
./mkfs.btrfs(main+0x1fe9)[0x564d1a7b442a]
/lib64/libc.so.6(__libc_start_main+0xcd)[0x7f36377d37fd]
./mkfs.btrfs(_start+0x2a)[0x564d1a7b1fda]
zsh: IOT instruction sudo ./mkfs.btrfs -f -d single -m single -R quota /dev/nullb0
The issue occurs because btrfs_create_root() is not formatting the root
node properly. This is fine in regular mode, because it's fortunately
reusing an once freed buffer. As the previous tree node allocation
kindly formatted the header, it will see the proper bytenr and pass the
checks.
However, we never reuse a once freed buffer on zoned filesystem. As a
result, we have zero-filled bytenr, FSID, and chunk-tree UUID, hitting
the asserts in check_tree_block().
Reported-by: Johannes Thumshirn <Johannes.Thumshirn@wdc.com>
Signed-off-by: Naohiro Aota <naohiro.aota@wdc.com>
Signed-off-by: David Sterba <dsterba@suse.com>
The btrfs_check_node() has far less meaningful error message compared to
kernel counterpart, and it even lacks certain checks like level check.
Backport btrfs_check_node() to btrfs-progs to not only unify the code
but greatly improve the readability of the error messages.
Extra modification includes:
- No fs_info needed
As we don't need to output fsid.
- Remove unlikely() macro
- Extra BTRFS_TREE_BLOCK_* error type
- Btrfs-progs specific error handling
To record the corrupted tree blocks.
Signed-off-by: Qu Wenruo <wqu@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Currently btrfs_check_leaf() provides almost meaningless messages for
things like invalid item offset:
incorrect offsets 8492 3707786077
While kernel tree-checker is doing a way better job, so it's wise to
backport btrfs_check_leaf() from kernel.
There are some modification needed:
- New generic_err() helper
- Remove unlikely() macro
- Remove empty essential tree check
Mkfs still needs to create empty essential trees.
- Using BTRFS_TREE_BLOCK_* return value
Original mode check still relies on them to do certain repair.
- No need for btrfs_fs_info
We no longer need fsid output, thus no need for btrfs_fs_info.
- No item contents check
- Still using the fail: label for btrfs-progs specific error handling
The new output looks like:
corrupt leaf: root=2 block=72164753408 slot=109, unexpected item end, have 3707786077 expect 8492
Signed-off-by: Qu Wenruo <wqu@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
In kernel space we hardly use btrfs_disk_key, unless for very lowlevel
code.
There is no need to intentionally use btrfs_disk_key in btrfs-progs
either.
Signed-off-by: Qu Wenruo <wqu@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Add the GPL v2 header to files where it was missing and is not from an
external source, update to the most recent version with the address.
Signed-off-by: David Sterba <dsterba@suse.com>
By enabling the lowmem checks properly I uncovered the case where test
fsck/007 will infinite loop at the detection stage. This is because
when checking the inode item we will just btrfs_next_item(), and because
we ignore check tree block failures at read time we don't get an -EIO
from btrfs_next_leaf.
This occurs because we allow fsck to raw-read blocks even if they fail
basic sanity checks, because we want the opportunity to repair the
blocks. However this means corrupt blocks are sitting in cache marked
as uptodate. btrfs_search_slot() handles this by doing a check_block()
on every block we add to the path, so that anything that is doing a
search gets a proper -EIO.
btrfs_next_sibling_block() needs a similar check. With this fix we now
return -EIO on btrfs_next_leaf() properly and we no longer infinite loop
on fsck/007 with lowmem.
Signed-off-by: Josef Bacik <josef@toxicpanda.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Add new options to dumps checksums in node headers and in the checksum
items:
$ btrfs inspect dump-tree --csum-headers image
root tree
leaf 471515136 items 19 free space 12186 generation 15 owner ROOT_TREE
leaf 471515136 flags 0x1(WRITTEN) backref revision 1 csum 0x756b2d54
fs uuid df0348df-5773-47dd-81e9-a18221461239
For nodes/leaves it's appended on the 2nd line of the header.
Checksum items are stored in leaves as EXTENT_CSUM key type, with offset
value as the logical offset starting. As the array would be hard to
parse or match, each offset value is printed with the checksum. For
crc32c it's 4 values on a line, for xxhash it's 2 and for the long
256bit checksums it's one checksum per line.
$ btrfs inspect dump-tree --csum-items image
leaf 5423104 items 1 free space 30 generation 6 owner CSUM_TREE
leaf 5423104 flags 0x1(WRITTEN) backref revision 1
fs uuid bd7c981e-16ff-4081-a734-3ef5d50cafc1
chunk uuid 13f4c76c-7845-4984-88ed-f01b52e05cf8
item 0 key (EXTENT_CSUM EXTENT_CSUM 22020096) itemoff 55 itemsize 16228
range start 22020096 end 38637568 length 16617472
[22020096] 0x8941f998 [22024192] 0x8941f998 [22028288] 0x8941f998 [22032384] 0x8941f998
[22036480] 0x8941f998 [22040576] 0x8941f998 [22044672] 0x8941f998 [22048768] 0x8941f998
...
$ btrfs inspect dump-tree --csum-items image
leaf 5718016 items 1 free space 7746 generation 6 owner CSUM_TREE
leaf 5718016 flags 0x1(WRITTEN) backref revision 1
fs uuid f453a5b4-8b4a-4fbf-90a2-2925e4fe2335
chunk uuid eb1da63b-248b-44c2-82da-71b2564bf50e
item 0 key (EXTENT_CSUM EXTENT_CSUM 52387840) itemoff 7771 itemsize 8512
range start 52387840 end 53477376 length 1089536
[52387840] 0x686ede9288c391e7e05026e56f2f91bfd879987a040ea98445dabc76f55b8e5f
[52391936] 0x686ede9288c391e7e05026e56f2f91bfd879987a040ea98445dabc76f55b8e5f
...
The options are not on by default, the header checksum is not important
for the structures. Data checksums can be quite big so that would make
the dump long and without any actual data to match against.
Signed-off-by: David Sterba <dsterba@suse.com>
Move the file to common as it's used by several parts, while still
keeping the name 'repair' although the only thing it does is adding a
corrupted extent.
Signed-off-by: David Sterba <dsterba@suse.com>
In case the right buffer is emptied it's first set to NULL and
subsequently it's dereferenced to get its size to pass to root_sub_used.
This naturally leads to a NULL pointer dereference. The correct thing to
do is to pass the stashed right->len in "blocksize".
Issue: #296
Pull-request: #360
Reviewed-by: Josef Bacik <josef@toxicpanda.com>
Signed-off-by: Nikolay Borisov <nborisov@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
