Instead of passing the transaction and subvolume root as arguments to
replay_one_extent(), pass the walk_control structure as we can grab all
of those from the structure. This reduces the number of arguments passed
and it's going to be needed by an incoming change that improves error
reporting for log replay.
Signed-off-by: Filipe Manana <fdmanana@suse.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Instead of passing the transaction and log tree as arguments to
check_item_in_log(), pass the walk_control structure as we can grab those
from the structure. This reduces the number of arguments passed and it's
going to be needed by an incoming change that improves error reporting for
log replay. Notice that a NULL log root argument to check_item_in_log()
makes it unconditionally delete a directory entry, so since the
walk_control always has a non-NULL log root, we add an extra boolean to
check_item_in_log() to tell it if it should unconditionally delete a
directory entry, preserving the behaviour and also making it a bit more
clear.
Signed-off-by: Filipe Manana <fdmanana@suse.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Instead of passing the transaction, subvolume root and log tree as
arguments to replay_dir_deletes(), pass the walk_control structure as
we can grab all of those from the structure. This reduces the number of
arguments passed and it's going to be needed by an incoming change that
improves error reporting for log replay. This also requires changing
fixup_inode_link_counts() and fixup_inode_link_count() to take that
structure as an argument since fixup_inode_link_count() makes a call
to replay_dir_deletes().
Signed-off-by: Filipe Manana <fdmanana@suse.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
In upcoming changes we need to pass struct walk_control as an argument to
replay_dir_deletes() and link_to_fixup_dir() so we need to move its
definition above the prototypes of those functions. So move it up right
below the enum that defines log replay stages and before any functions and
function prototypes are declared. Also fixup the comments while moving it
so that they comply with the preferred code style (capitalize the first
word in a sentence, end sentences with punctuation, makes lines wider and
closer to the 80 characters limit).
Signed-off-by: Filipe Manana <fdmanana@suse.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Instead of passing the transaction, subvolume root and log tree as
arguments to replay_xattr_deletes(), pass the walk_control structure as
we can grab all of those from the structure. This reduces the number of
arguments passed and it's going to be needed by an incoming change that
improves error reporting for log replay.
Signed-off-by: Filipe Manana <fdmanana@suse.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Currently we have this odd behaviour:
1) At btrfs_replay_log() we drop the reference of the log root tree if
the call to btrfs_recover_log_trees() failed;
2) But if the call to btrfs_recover_log_trees() did not fail, we don't
drop the reference in btrfs_replay_log() - we expect that
btrfs_recover_log_trees() does it in case it returns success.
Let's simplify this and make btrfs_replay_log() always drop the reference
on the log root tree, not only this simplifies code as it's what makes
sense since it's btrfs_replay_log() who grabbed the reference in the first
place.
Signed-off-by: Filipe Manana <fdmanana@suse.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
There's no point in setting log_root_tree->log_root to NULL as this is
already NULL, we never assigned anything to it before and it's meaningless
as a log root never has a value other than NULL for the ->log_root field,
that can be not NULL only for non log roots.
Signed-off-by: Filipe Manana <fdmanana@suse.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
It's unncessary to pass a transaction parameter since struct walk_control
already has a member that points to the transaction, so we can make the
functions access the structure.
Signed-off-by: Filipe Manana <fdmanana@suse.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Instead of duplicating the dropping of a log tree in case we jump to the
'error' label, move the dropping under the 'error' label and get rid of the
the unnecessary setting of the log root to NULL since we return immediately
after.
Signed-off-by: Filipe Manana <fdmanana@suse.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Instead of passing an extra log root parameter for the log tree walk
functions and callbacks, add the log tree to struct walk_control and
make those functions and callbacks extract the log root from that
structure, reducing the number of parameters. This also simplifies
further upcoming changes to report log tree replay failures.
Signed-off-by: Filipe Manana <fdmanana@suse.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Everywhere we have a log root we name it as 'log' or 'log_root' except in
walk_down_log_tree() and walk_up_log_tree() where we name it as 'root',
which not only it's inconsistent, it's also confusing since we typically
use 'root' when naming variables that refer to a subvolume tree. So for
clairty and consistency rename the 'root' argument to 'log'.
Signed-off-by: Filipe Manana <fdmanana@suse.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Everywhere else we refer to a subvolume root we are replaying to simply
as 'root', so rename from 'replay_dest' to 'root' for consistency and
having a more meaningful and shorter name. While at it also update the
comment to be more detailed and comply to preferred style (first word in
a sentence is capitalized and sentence ends with punctuation).
Signed-off-by: Filipe Manana <fdmanana@suse.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
The 'free' and 'pin' member of struct walk_control, used during log replay
and when freeing a log tree, are defined as integers but in practice are
used as booleans. Change their type to bool and while at it update their
comments to be more detailed and comply with the preferred comment style
(first word in a sentence is capitalized, sentences end with punctuation
and the comment opening (/*) is on a line of its own).
Signed-off-by: Filipe Manana <fdmanana@suse.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Inside btrfs_fs_info we cache several bits shift like sectorsize_bits.
Apply this to max and min folio orders so that every time mapping order
needs to be applied we can skip the calculation.
Signed-off-by: Qu Wenruo <wqu@suse.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Currently if we want to iterate all blocks inside a bio, we do something
like this:
bio_for_each_segment_all(bvec, bio, iter_all) {
for (off = 0; off < bvec->bv_len; off += sectorsize) {
/* Iterate blocks using bv + off */
}
}
That's fine for now, but it will not handle future bs > ps, as
bio_for_each_segment_all() is a single-page iterator, it will always
return a bvec that's no larger than a page.
But for bs > ps cases, we need a full folio (which covers at least one
block) so that we can work on the block.
To address this problem and handle future bs > ps cases better:
- Introduce a helper btrfs_bio_for_each_block_all()
This helper will create a local bvec_iter, which has the size of the
target bio. Then grab the current physical address of the current
location, then advance the iterator by block size.
- Use btrfs_bio_for_each_block_all() to replace existing call sites
Including:
* set_bio_pages_uptodate() in raid56
* verify_bio_data_sectors() in raid56
Both will result much easier to read code.
Signed-off-by: Qu Wenruo <wqu@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Currently if we want to iterate a bio in block unit, we do something
like this:
while (iter->bi_size) {
struct bio_vec bv = bio_iter_iovec();
/* Do something with using the bv */
bio_advance_iter_single(&bbio->bio, iter, sectorsize);
}
That's fine for now, but it will not handle future bs > ps, as
bio_iter_iovec() returns a single-page bvec, meaning the bv_len will not
exceed page size.
This means the code using that bv can only handle a block if bs <= ps.
To address this problem and handle future bs > ps cases better:
- Introduce a helper btrfs_bio_for_each_block()
Instead of bio_vec, which has single and multiple page version and
multiple page version has quite some limits, use my favorite way to
represent a block, phys_addr_t.
For bs <= ps cases, nothing is changed, except we will do a very
small overhead to convert phys_addr_t to a folio, then use the proper
folio helpers to handle the possible highmem cases.
For bs > ps cases, all blocks will be backed by large folios, meaning
every folio will cover at least one block. And still use proper folio
helpers to handle highmem cases.
With phys_addr_t, we will handle both large folio and highmem
properly. So there is no better single variable to present a btrfs
block than phys_addr_t.
- Extract the data block csum calculation into a helper
The new helper, btrfs_calculate_block_csum() will be utilized by
btrfs_csum_one_bio().
- Use btrfs_bio_for_each_block() to replace existing call sites
Including:
* index_one_bio() from raid56.c
Very straight-forward.
* btrfs_check_read_bio()
Also update repair_one_sector() to grab the folio using phys_addr_t,
and do extra checks to make sure the folio covers at least one
block.
We do not need to bother bv_len at all now.
* btrfs_csum_one_bio()
Now we can move the highmem handling into a dedicated helper,
calculate_block_csum(), and use btrfs_bio_for_each_block() helper.
There is one exception in btrfs_decompress_buf2page(), which is copying
decompressed data into the original bio, which is not iterating using
block size thus we don't need to bother.
Signed-off-by: Qu Wenruo <wqu@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Currently for btrfs checksum verification, we do it in the following
pattern:
kaddr = kmap_local_*();
ret = btrfs_check_csum_csum(kaddr);
kunmap_local(kaddr);
It's OK for now, but it's still not following the patterns of helpers
inside linux/highmem.h, which never requires a virt memory address.
In those highmem helpers, they mostly accept a folio, some offset/length
inside the folio, and in the implementation they check if the folio
needs partial kmap, and do the handling.
Inspired by those formal highmem helpers, enhance the highmem handling
of data checksum verification by:
- Rename btrfs_check_sector_csum() to btrfs_check_block_csum()
To follow the more common term "block" used in all other major
filesystems.
- Pass a physical address into btrfs_check_block_csum() and
btrfs_data_csum_ok()
The physical address is always available even for a highmem page.
Since it's page frame number << PAGE_SHIFT + offset in page.
And with that physical address, we can grab the folio covering the
page, and do extra checks to ensure it covers at least one block.
This also allows us to do the kmap inside btrfs_check_block_csum().
This means all the extra HIGHMEM handling will be concentrated into
btrfs_check_block_csum(), and no callers will need to bother highmem
by themselves.
- Properly zero out the block if csum mismatch
Since btrfs_data_csum_ok() only got a paddr, we can not and should not
use memzero_bvec(), which only accepts single page bvec.
Instead use paddr to grab the folio and call folio_zero_range()
Signed-off-by: Qu Wenruo <wqu@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Currently if block size < page size, btrfs only supports one single
config, 4K.
This is mostly to reduce the test configurations, as 4K is going to be
the default block size for all architectures.
However all other major filesystems have no artificial limits on the
support block size, and some are already supporting block size > page
sizes.
Since the btrfs subpage block support has been there for a long time,
it's time for us to enable all block size <= page size support.
So here enable all block sizes support as long as it's no larger than
page size for experimental builds.
Signed-off-by: Qu Wenruo <wqu@suse.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Replace max_t() followed by min_t() with a single clamp().
As was pointed by David Laight in
https://lore.kernel.org/linux-btrfs/20250906122458.75dfc8f0@pumpkin/
the calculation may overflow u32 when the input value is too large, so
clamp_t() is not used. In practice the expected values are in range of
megabytes to gigabytes (throughput limit) so the bug would not happen.
Signed-off-by: Thorsten Blum <thorsten.blum@linux.dev>
Reviewed-by: David Sterba <dsterba@suse.com>
[ Use clamp() and add explanation. ]
Signed-off-by: David Sterba <dsterba@suse.com>
Annual typo fixing pass. Strangely codespell found only about 30% of
what is in this patch, the rest was done manually using text
spellchecker with a custom dictionary of acceptable terms.
Reviewed-by: Neal Gompa <neal@gompa.dev>
Signed-off-by: David Sterba <dsterba@suse.com>
Currently the compression workspace buffer size is always based on
PAGE_SIZE, but btrfs has support subpage sized block size for some time.
This means for one-shot compression algorithm like lzo, we're wasting
quite some memory if the block size is smaller than page size, as the
LZO only works on one block (thus one-shot).
On 64K page sized systems with 4K block size, it means we only need at
most 8K buffer space for lzo, but in reality we're allocating 64K
buffer.
So to reduce the memory usage, change all workspace buffer to base its
size based on block size other than page size.
Signed-off-by: Qu Wenruo <wqu@suse.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Since all workspace managers are per-fs, there is no need nor no way to
store them inside btrfs_compress_op::wsm anymore.
With that said, we can do the following modifications:
- Remove zstd_workspace_mananger::ops
Zstd always grab the global btrfs_compress_op[].
- Remove btrfs_compress_op::wsm member
- Rename btrfs_compress_op to btrfs_compress_levels
This should make it more clear that btrfs_compress_levels structures are
only to indicate the levels of each compress algorithm.
Signed-off-by: Qu Wenruo <wqu@suse.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Since all workspaces are handled by the per-fs workspace managers, we
can safely remove the old per-module managers.
Signed-off-by: Qu Wenruo <wqu@suse.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
There are several interfaces involved for each algorithm:
- alloc workspace
All algorithms allocate a workspace without the need for workspace
manager.
So no change needs to be done.
- get workspace
This involves checking the workspace manager to find a free one, and
if not, allocate a new one.
For none and lzo, they share the same generic btrfs_get_workspace()
helper, only needs to update that function to use the per-fs manager.
For zlib it uses a wrapper around btrfs_get_workspace(), so no special
work needed.
For zstd, update zstd_find_workspace() and zstd_get_workspace() to
utilize the per-fs manager.
- put workspace
For none/zlib/lzo they share the same btrfs_put_workspace(), update
that function to use the per-fs manager.
For zstd, it's zstd_put_workspace(), the same update.
- zstd specific timer
This is the timer to reclaim workspace, change it to grab the per-fs
workspace manager instead.
Now all workspace are managed by the per-fs manager.
Signed-off-by: Qu Wenruo <wqu@suse.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
This involves:
- Add (alloc|free)_workspace_manager helpers.
These are the helper to alloc/free workspace_manager structure.
The allocator will allocate a workspace_manager structure, initialize
it, and pre-allocate one workspace for it.
The freer will do the cleanup and set the manager pointer to NULL.
- Call alloc_workspace_manager() inside btrfs_alloc_compress_wsm()
- Call alloc_workspace_manager() inside btrfs_free_compress_wsm()
For none, zlib and lzo compression algorithms.
For now the generic per-fs workspace managers won't really have any effect,
and all compression is still going through the global workspace manager.
Signed-off-by: Qu Wenruo <wqu@suse.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
This involves:
- Add zstd_alloc_workspace_manager() and zstd_free_workspace_manager()
Those two functions will accept an fs_info pointer, and alloc/free
fs_info->compr_wsm[BTRFS_COMPRESS_ZSTD] pointer.
- Add btrfs_alloc_compress_wsm() and btrfs_free_compress_wsm()
Those are helpers allocating the workspace managers for all
algorithms.
For now only zstd is supported, and the timing is a little unusual,
the btrfs_alloc_compress_wsm() should only be called after the
sectorsize being initialized.
Meanwhile btrfs_free_fs_info_compress() is called in
btrfs_free_fs_info().
- Move the definition of btrfs_compression_type to "fs.h"
The reason is that "compression.h" has already included "fs.h", thus
we can not just include "compression.h" to get the definition of
BTRFS_NR_COMPRESS_TYPES to define fs_info::compr_wsm[].
For now the per-fs zstd workspace manager won't really have any effect,
and all compression is still going through the global workspace manager.
Signed-off-by: Qu Wenruo <wqu@suse.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
[BACKGROUND]
Currently btrfs shares workspaces and their managers for all filesystems,
this is mostly fine as all those workspaces are using page size based
buffers, and btrfs only support block size (bs) <= page size (ps).
This means even if bs < ps, we at most waste some buffer space in the
workspace, but everything will still work fine.
The problem here is that is limiting our support for bs > ps cases.
As now a workspace now may need larger buffer to handle bs > ps cases,
but since the pool has no way to distinguish different workspaces, a
regular workspace (which is still using buffer size based on ps) can be
passed to a btrfs whose bs > ps.
In that case the buffer is not large enough, and will cause various
problems.
[ENHANCEMENT]
To prepare for the per-fs workspace migration, add an fs_info parameter
to all workspace related functions.
For now this new fs_info parameter is not yet utilized.
Signed-off-by: Qu Wenruo <wqu@suse.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
[BUG]
There is a very low chance that DEBUG_WARN() inside
btrfs_writepage_cow_fixup() can be triggered when
CONFIG_BTRFS_EXPERIMENTAL is enabled.
This only happens after run_delalloc_nocow() failed.
Unfortunately I haven't hit it for a while thus no real world dmesg for
now.
[CAUSE]
There is a race window where after run_delalloc_nocow() failed, error
handling can race with writeback thread.
Before we hit run_delalloc_nocow(), there is an inode with the following
dirty pages: (4K page size, 4K block size, no large folio)
0 4K 8K 12K 16K
|/////////|///////////|///////////|////////////|
The inode also have NODATACOW flag, and the above dirty range will go
through different extents during run_delalloc_range():
0 4K 8K 12K 16K
| NOCOW | COW | COW | NOCOW |
The race happen like this:
writeback thread A | writeback thread B
----------------------------------+--------------------------------------
Writeback for folio 0 |
run_delalloc_nocow() |
|- nocow_one_range() |
| For range [0, 4K), ret = 0 |
| |
|- fallback_to_cow() |
| For range [4K, 8K), ret = 0 |
| Folio 4K *UNLOCKED* |
| | Writeback for folio 4K
|- fallback_to_cow() | extent_writepage()
| For range [8K, 12K), failure | |- writepage_delalloc()
| | |
|- btrfs_cleanup_ordered_extents()| |
|- btrfs_folio_clear_ordered() | |
| Folio 0 still locked, safe | |
| | | Ordered extent already allocated.
| | | Nothing to do.
| | |- extent_writepage_io()
| | |- btrfs_writepage_cow_fixup()
|- btrfs_folio_clear_ordered() | |
Folio 4K hold by thread B, | |
UNSAFE! | |- btrfs_test_ordered()
| | Cleared by thread A,
| |
| |- DEBUG_WARN();
This is only possible after run_delalloc_nocow() failure, as
cow_file_range() will keep all folios and io tree range locked, until
everything is finished or after error handling.
The root cause is we allow fallback_to_cow() and nocow_one_range() to
unlock the folios after a successful run, so that during error handling
we're no longer safe to use btrfs_cleanup_ordered_extents() as the
folios are already unlocked.
[FIX]
- Make fallback_to_cow() and nocow_one_range() to keep folios locked
after a successful run
For fallback_to_cow() we can pass COW_FILE_RANGE_KEEP_LOCKED flag
into cow_file_range().
For nocow_one_range() we have to remove the PAGE_UNLOCK flag from
extent_clear_unlock_delalloc().
- Unlock folios if everything is fine in run_delalloc_nocow()
- Use extent_clear_unlock_delalloc() to handle range [@start,
@cur_offset) inside run_delalloc_nocow()
Since folios are still locked, we do not need
cleanup_dirty_folios() to do the cleanup.
extent_clear_unlock_delalloc() with "PAGE_START_WRITEBACK |
PAGE_END_WRITEBACK" will clear the dirty flags.
- Remove cleanup_dirty_folios()
Signed-off-by: Qu Wenruo <wqu@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Currently if we hit an error inside nocow_one_range(), we do not clear
the page dirty, and let the caller to handle it.
This is very different compared to fallback_to_cow(), when that function
failed, everything will be cleaned up by cow_file_range().
Enhance the situation by:
- Use a common error handling for nocow_one_range()
If we failed anything, use the same btrfs_cleanup_ordered_extents()
and extent_clear_unlock_delalloc().
btrfs_cleanup_ordered_extents() is safe even if we haven't created new
ordered extent, in that case there should be no OE and that function
will do nothing.
The same applies to extent_clear_unlock_delalloc(), and since we're
passing PAGE_UNLOCK | PAGE_START_WRITEBACK | PAGE_END_WRITEBACK, it
will also clear folio dirty flag during error handling.
- Avoid touching the failed range of nocow_one_range()
As the failed range will be cleaned up and unlocked by that function.
Here we introduce a new variable @nocow_end to record the failed range,
so that we can skip it during the error handling of run_delalloc_nocow().
Signed-off-by: Qu Wenruo <wqu@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
When running emulated write error tests like generic/475, we can hit
error messages like this:
BTRFS error (device dm-12 state EA): run_delalloc_nocow failed, root=596 inode=264 start=1605632 len=73728: -5
BTRFS error (device dm-12 state EA): failed to run delalloc range, root=596 ino=264 folio=1605632 submit_bitmap=0-7 start=1605632 len=73728: -5
Which is normally buried by direct IO error messages.
However above error messages are not enough to determine which is the
real range that caused the error.
Considering we can have multiple different extents in one delalloc
range (e.g. some COW extents along with some NOCOW extents), just
outputting the error at the end of run_delalloc_nocow() is not enough.
To enhance the error messages:
- Remove the rate limit on the existing error messages
In the generic/475 example, most error messages are from direct IO,
not really from the delalloc range.
Considering how useful the delalloc range error messages are, we don't
want they to be rate limited.
- Add extra @cur_offset output for cow_file_range()
- Add extra variable output for run_delalloc_nocow()
This is especially important for run_delalloc_nocow(), as there
are extra error paths where we can hit error without into
nocow_one_range() nor fallback_to_cow().
- Add an error message for nocow_one_range()
That's the missing part.
For fallback_to_cow(), we have error message from cow_file_range()
already.
- Constify the @len and @end local variables for nocow_one_range()
This makes it much easier to make sure @len and @end are not modified
at runtime.
Reviewed-by: Boris Burkov <boris@bur.io>
Signed-off-by: Qu Wenruo <wqu@suse.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Currently the error handling of run_delalloc_nocow() modifies
@cur_offset to handle different parts of the delalloc range.
However the error handling can always be split into 3 parts:
1) The range with ordered extents allocated (OE cleanup)
We have to cleanup the ordered extents and unlock the folios.
2) The range that have been cleaned up (skip)
For now it's only for the range of fallback_to_cow().
We should not touch it at all.
3) The range that is not yet touched (untouched)
We have to unlock the folios and clear any reserved space.
This 3 ranges split has the same principle as cow_file_range(), however
the NOCOW/COW handling makes the above 3 range split much more complex:
a) Failure immediately after a successful OE allocation
Thus no @cow_start nor @cow_end set.
start cur_offset end
| OE cleanup | untouched |
b) Failure after hitting a COW range but before calling
fallback_to_cow()
start cow_start cur_offset end
| OE Cleanup | untouched |
c) Failure to call fallback_to_cow()
start cow_start cow_end end
| OE Cleanup | skip | untouched |
Instead of modifying @cur_offset, do proper range calculation for
OE-cleanup and untouched ranges using above 3 cases with proper range
charts.
This avoid updating @cur_offset, as it will an extra output for debug
purposes later, and explain the behavior easier.
Signed-off-by: Qu Wenruo <wqu@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
The ref_tracker infrastructure aids debugging but is not enabled by
default as it has a performance impact. Add mount option 'ref_tracker'
so it can be selectively enabled on a filesystem. Currently it track
references of 'delayed inodes'.
Signed-off-by: Leo Martins <loemra.dev@gmail.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
We are seeing soft lockups in kill_all_delayed_nodes due to a
delayed_node with a lingering reference count of 1. Printing at this
point will reveal the guilty stack trace. If the delayed_node has no
references there should be no output.
Signed-off-by: Leo Martins <loemra.dev@gmail.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Add ref_tracker infrastructure for struct btrfs_delayed_node.
It is a response to the largest btrfs related crash in our fleet. We're
seeing soft lockups in btrfs_kill_all_delayed_nodes() that seem to be a
result of delayed_nodes not being released properly.
A ref_tracker object is allocated on reference count increases and freed
on reference count decreases. The ref_tracker object stores a stack
trace of where it is allocated. The ref_tracker_dir object is embedded
in btrfs_delayed_node and keeps track of all current and some old/freed
ref_tracker objects. When a leak is detected we can print the stack
traces for all ref_trackers that have not yet been freed.
Here is a common example of taking a reference to a delayed_node and
freeing it with ref_tracker.
struct btrfs_ref_tracker tracker;
struct btrfs_delayed_node *node;
node = btrfs_get_delayed_node(inode, &tracker);
// use delayed_node...
btrfs_release_delayed_node(node, &tracker);
There are two special cases where the delayed_node reference is "long
lived", meaning that the thread that takes the reference and the thread
that releases the reference are different. The 'inode_cache_tracker'
tracks the delayed_node stored in btrfs_inode. The 'node_list_tracker'
tracks the delayed_node stored in the btrfs_delayed_root
node_list/prepare_list. These trackers are embedded in the
btrfs_delayed_node.
btrfs_ref_tracker and btrfs_ref_tracker_dir are wrappers that either
compile to the corresponding ref_tracker structs or empty structs
depending on CONFIG_BTRFS_DEBUG. There are also btrfs wrappers for
the ref_tracker API.
Signed-off-by: Leo Martins <loemra.dev@gmail.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
We're almost done cleaning misused int/bool parameters. Convert a bunch
of them, found by manual grepping. Note that btrfs_sync_fs() needs an
int as it's mandated by the struct super_operations prototype.
Reviewed-by: Boris Burkov <boris@bur.io>
Signed-off-by: David Sterba <dsterba@suse.com>
Remove CONFIG_BTRFS_FS_REF_VERIFY Kconfig and add it as part of
CONFIG_BTRFS_DEBUG. This should not be impactful to the performance
of debug. The struct btrfs_ref takes an additional u64, btrfs_fs_info
takes an additional spinlock_t and rb_root. All of the ref_verify logic
is still protected by a mount option.
Signed-off-by: Leo Martins <loemra.dev@gmail.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Use the standard error pointer macro to simplify the code.
Reviewed-by: Daniel Vacek <neelx@suse.com>
Signed-off-by: Xichao Zhao <zhao.xichao@vivo.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Currently we manually check the block size against 3 different values:
- 4K
- PAGE_SIZE
- MIN_BLOCKSIZE
Those 3 values can match or differ from each other. This makes it
pretty complex to output the supported block sizes.
Considering we're going to add block size > page size support soon, this
can make the support block size sysfs attribute much harder to
implement.
To make it easier, factor out a helper, btrfs_supported_blocksize() to
do a simple check for the block size.
Then utilize it in the two locations:
- btrfs_validate_super()
This is very straightforward
- supported_sectorsizes_show()
Iterate through all valid block sizes, and only output supported ones.
This is to make future full range block sizes support much easier.
Reviewed-by: Anand Jain <anand.jain@oracle.com>
Signed-off-by: Qu Wenruo <wqu@suse.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
[BEHAVIOR DIFFERENCE BETWEEN COMPRESSION ALGOS]
Currently LZO compression algorithm will check if we're making the
compressed data larger after compressing more than 2 blocks.
But zlib and zstd do the same checks after compressing more than 8192
bytes.
This is not a big deal, but since we're already supporting larger block
size (e.g. 64K block size if page size is also 64K), this check is not
suitable for all block sizes.
For example, if our page and block size are both 16KiB, and after the
first block compressed using zlib, the resulted compressed data is
slightly larger than 16KiB, we will immediately abort the compression.
This makes zstd and zlib compression algorithms to behave slightly
different from LZO, which only aborts after compressing two blocks.
[ENHANCEMENT]
To unify the behavior, only abort the compression after compressing at
least two blocks.
Reviewed-by: Anand Jain <anand.jain@oracle.com>
Signed-off-by: Qu Wenruo <wqu@suse.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
For the 3 supported compression algorithms, two of them (zstd and zlib)
are already grabbing the btrfs inode for error messages.
It's more common to pass btrfs_inode and grab the address space from it.
Reviewed-by: Anand Jain <anand.jain@oracle.com>
Signed-off-by: Qu Wenruo <wqu@suse.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
The hint block group selection in the extent allocator is wrong in the
first place, as it can select the dedicated data relocation block group for
the normal data allocation.
Since we separated the normal data space_info and the data relocation
space_info, we can easily identify a block group is for data relocation or
not. Do not choose it for the normal data allocation.
Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com>
Signed-off-by: Naohiro Aota <naohiro.aota@wdc.com>
Signed-off-by: David Sterba <dsterba@suse.com>
If you run a workload with:
- a cgroup that does tons of parallel data reading, with a working set
much larger than its memory limit
- a second cgroup that writes relatively fewer files, with overwrites,
with no memory limit
(see full code listing at the bottom for a reproducer)
Then what quickly occurs is:
- we have a large number of threads trying to read the csum tree
- we have a decent number of threads deleting csums running delayed refs
- we have a large number of threads in direct reclaim and thus high
memory pressure
The result of this is that we writeback the csum tree repeatedly mid
transaction, to get back the extent_buffer folios for reclaim. As a
result, we repeatedly COW the csum tree for the delayed refs that are
deleting csums. This means repeatedly write locking the higher levels of
the tree.
As a result of this, we achieve an unpleasant priority inversion. We
have:
- a high degree of contention on the csum root node (and other upper
nodes) eb rwsem
- a memory starved cgroup doing tons of reclaim on CPU.
- many reader threads in the memory starved cgroup "holding" the sem
as readers, but not scheduling promptly. i.e., task __state == 0, but
not running on a cpu.
- btrfs_commit_transaction stuck trying to acquire the sem as a writer.
(running delayed_refs, deleting csums for unreferenced data extents)
This results in arbitrarily long transactions. This then results in
seriously degraded performance for any cgroup using the filesystem (the
victim cgroup in the script).
It isn't an academic problem, as we see this exact problem in production
at Meta with one cgroup over its memory limit ruining btrfs performance
for the whole system, stalling critical system services that depend on
btrfs syncs.
The underlying scheduling "problem" with global rwsems is sort of thorny
and apparently well known and was discussed at LPC 2024, for example.
As a result, our main lever in the short term is just trying to reduce
contention on our various rwsems with an eye to reducing the frequency
of write locking, to avoid disabling the read lock fast acquisition path.
Luckily, it seems likely that many reads are for old extents written
many transactions ago, and that for those we *can* in fact search the
commit root. The commit_root_sem only gets taken write once, near the
end of transaction commit, no matter how much memory pressure there is,
so we have much less contention between readers and writers.
This change detects when we are trying to read an old extent (according
to extent map generation) and then wires that through bio_ctrl to the
btrfs_bio, which unfortunately isn't allocated yet when we have this
information. When we go to lookup the csums in lookup_bio_sums we can
check this condition on the btrfs_bio and do the commit root lookup
accordingly.
Note that a single bio_ctrl might collect a few extent_maps into a single
bio, so it is important to track a maximum generation across all the
extent_maps used for each bio to make an accurate decision on whether it
is valid to look in the commit root. If any extent_map is updated in the
current generation, we can't use the commit root.
To test and reproduce this issue, I used the following script and
accompanying C program (to avoid bottlenecks in constantly forking
thousands of dd processes):
====== big-read.c ======
#include <fcntl.h>
#include <stdio.h>
#include <stdlib.h>
#include <sys/mman.h>
#include <sys/stat.h>
#include <unistd.h>
#include <errno.h>
#define BUF_SZ (128 * (1 << 10UL))
int read_once(int fd, size_t sz) {
char buf[BUF_SZ];
size_t rd = 0;
int ret = 0;
while (rd < sz) {
ret = read(fd, buf, BUF_SZ);
if (ret < 0) {
if (errno == EINTR)
continue;
fprintf(stderr, "read failed: %d\n", errno);
return -errno;
} else if (ret == 0) {
break;
} else {
rd += ret;
}
}
return rd;
}
int read_loop(char *fname) {
int fd;
struct stat st;
size_t sz = 0;
int ret;
while (1) {
fd = open(fname, O_RDONLY);
if (fd == -1) {
perror("open");
return 1;
}
if (!sz) {
if (!fstat(fd, &st)) {
sz = st.st_size;
} else {
perror("stat");
return 1;
}
}
ret = read_once(fd, sz);
close(fd);
}
}
int main(int argc, char *argv[]) {
int fd;
struct stat st;
off_t sz;
char *buf;
int ret;
if (argc != 2) {
fprintf(stderr, "Usage: %s <filename>\n", argv[0]);
return 1;
}
return read_loop(argv[1]);
}
====== repro.sh ======
#!/usr/bin/env bash
SCRIPT=$(readlink -f "$0")
DIR=$(dirname "$SCRIPT")
dev=$1
mnt=$2
shift
shift
CG_ROOT=/sys/fs/cgroup
BAD_CG=$CG_ROOT/bad-nbr
GOOD_CG=$CG_ROOT/good-nbr
NR_BIGGOS=1
NR_LITTLE=10
NR_VICTIMS=32
NR_VILLAINS=512
START_SEC=$(date +%s)
_elapsed() {
echo "elapsed: $(($(date +%s) - $START_SEC))"
}
_stats() {
local sysfs=/sys/fs/btrfs/$(findmnt -no UUID $dev)
echo "================"
date
_elapsed
cat $sysfs/commit_stats
cat $BAD_CG/memory.pressure
}
_setup_cgs() {
echo "+memory +cpuset" > $CG_ROOT/cgroup.subtree_control
mkdir -p $GOOD_CG
mkdir -p $BAD_CG
echo max > $BAD_CG/memory.max
# memory.high much less than the working set will cause heavy reclaim
echo $((1 << 30)) > $BAD_CG/memory.high
# victims get a subset of villain CPUs
echo 0 > $GOOD_CG/cpuset.cpus
echo 0,1,2,3 > $BAD_CG/cpuset.cpus
}
_kill_cg() {
local cg=$1
local attempts=0
echo "kill cgroup $cg"
[ -f $cg/cgroup.procs ] || return
while true; do
attempts=$((attempts + 1))
echo 1 > $cg/cgroup.kill
sleep 1
procs=$(wc -l $cg/cgroup.procs | cut -d' ' -f1)
[ $procs -eq 0 ] && break
done
rmdir $cg
echo "killed cgroup $cg in $attempts attempts"
}
_biggo_vol() {
echo $mnt/biggo_vol.$1
}
_biggo_file() {
echo $(_biggo_vol $1)/biggo
}
_subvoled_biggos() {
total_sz=$((10 << 30))
per_sz=$((total_sz / $NR_VILLAINS))
dd_count=$((per_sz >> 20))
echo "create $NR_VILLAINS subvols with a file of size $per_sz bytes for a total of $total_sz bytes."
for i in $(seq $NR_VILLAINS)
do
btrfs subvol create $(_biggo_vol $i) &>/dev/null
dd if=/dev/zero of=$(_biggo_file $i) bs=1M count=$dd_count &>/dev/null
done
echo "done creating subvols."
}
_setup() {
[ -f .done ] && rm .done
findmnt -n $dev && exit 1
if [ -f .re-mkfs ]; then
mkfs.btrfs -f -m single -d single $dev >/dev/null || exit 2
else
echo "touch .re-mkfs to populate the test fs"
fi
mount -o noatime $dev $mnt || exit 3
[ -f .re-mkfs ] && _subvoled_biggos
_setup_cgs
}
_my_cleanup() {
echo "CLEANUP!"
_kill_cg $BAD_CG
_kill_cg $GOOD_CG
sleep 1
umount $mnt
}
_bad_exit() {
_err "Unexpected Exit! $?"
_stats
exit $?
}
trap _my_cleanup EXIT
trap _bad_exit INT TERM
_setup
# Use a lot of page cache reading the big file
_villain() {
local i=$1
echo $BASHPID > $BAD_CG/cgroup.procs
$DIR/big-read $(_biggo_file $i)
}
# Hit del_csum a lot by overwriting lots of small new files
_victim() {
echo $BASHPID > $GOOD_CG/cgroup.procs
i=0;
while (true)
do
local tmp=$mnt/tmp.$i
dd if=/dev/zero of=$tmp bs=4k count=2 >/dev/null 2>&1
i=$((i+1))
[ $i -eq $NR_LITTLE ] && i=0
done
}
_one_sync() {
echo "sync..."
before=$(date +%s)
sync
after=$(date +%s)
echo "sync done in $((after - before))s"
_stats
}
# sync in a loop
_sync() {
echo "start sync loop"
syncs=0
echo $BASHPID > $GOOD_CG/cgroup.procs
while true
do
[ -f .done ] && break
_one_sync
syncs=$((syncs + 1))
[ -f .done ] && break
sleep 10
done
if [ $syncs -eq 0 ]; then
echo "do at least one sync!"
_one_sync
fi
echo "sync loop done."
}
_sleep() {
local time=${1-60}
local now=$(date +%s)
local end=$((now + time))
while [ $now -lt $end ];
do
echo "SLEEP: $((end - now))s left. Sleep 10."
sleep 10
now=$(date +%s)
done
}
echo "start $NR_VILLAINS villains"
for i in $(seq $NR_VILLAINS)
do
_villain $i &
disown # get rid of annoying log on kill (done via cgroup anyway)
done
echo "start $NR_VICTIMS victims"
for i in $(seq $NR_VICTIMS)
do
_victim &
disown
done
_sync &
SYNC_PID=$!
_sleep $1
_elapsed
touch .done
wait $SYNC_PID
echo "OK"
exit 0
Without this patch, that reproducer:
- Ran for 6+ minutes instead of 60s
- Hung hundreds of threads in D state on the csum reader lock
- Got a commit stuck for 3 minutes
sync done in 388s
================
Wed Jul 9 09:52:31 PM UTC 2025
elapsed: 420
commits 2
cur_commit_ms 0
last_commit_ms 159446
max_commit_ms 159446
total_commit_ms 160058
some avg10=99.03 avg60=98.97 avg300=75.43 total=418033386
full avg10=82.79 avg60=80.52 avg300=59.45 total=324995274
419 hits state R, D comms big-read
btrfs_tree_read_lock_nested
btrfs_read_lock_root_node
btrfs_search_slot
btrfs_lookup_csum
btrfs_lookup_bio_sums
btrfs_submit_bbio
1 hits state D comms btrfs-transacti
btrfs_tree_lock_nested
btrfs_lock_root_node
btrfs_search_slot
btrfs_del_csums
__btrfs_run_delayed_refs
btrfs_run_delayed_refs
With the patch, the reproducer exits naturally, in 65s, completing a
pretty decent 4 commits, despite heavy memory pressure. Occasionally you
can still trigger a rather long commit (couple seconds) but never one
that is minutes long.
sync done in 3s
================
elapsed: 65
commits 4
cur_commit_ms 0
last_commit_ms 485
max_commit_ms 689
total_commit_ms 2453
some avg10=98.28 avg60=64.54 avg300=19.39 total=64849893
full avg10=74.43 avg60=48.50 avg300=14.53 total=48665168
some random rwalker samples showed the most common stack in reclaim,
rather than the csum tree:
145 hits state R comms bash, sleep, dd, shuf
shrink_folio_list
shrink_lruvec
shrink_node
do_try_to_free_pages
try_to_free_mem_cgroup_pages
reclaim_high
Link: https://lpc.events/event/18/contributions/1883/
Reviewed-by: Filipe Manana <fdmanana@suse.com>
Signed-off-by: Boris Burkov <boris@bur.io>
Signed-off-by: David Sterba <dsterba@suse.com>
In messages.h there's linux/types.h included more than once.
Reported-by: Abaci Robot <abaci@linux.alibaba.com>
Closes: https://bugzilla.openanolis.cn/show_bug.cgi?id=22939
Signed-off-by: Jiapeng Chong <jiapeng.chong@linux.alibaba.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Now that btrfs_zone_finish_endio_workfn() is directly calling
do_zone_finish() the only caller of btrfs_zone_finish_endio() is
btrfs_finish_one_ordered().
btrfs_finish_one_ordered() already has error handling in-place so
btrfs_zone_finish_endio() can return an error if the block group lookup
fails.
Also as btrfs_zone_finish_endio() already checks for zoned filesystems and
returns early, there's no need to do this in the caller.
Reviewed-by: Damien Le Moal <dlemoal@kernel.org>
Signed-off-by: Johannes Thumshirn <johannes.thumshirn@wdc.com>
Signed-off-by: David Sterba <dsterba@suse.com>
When btrfs_zone_finish_endio_workfn() is calling btrfs_zone_finish_endio()
it already has a pointer to the block group. Furthermore
btrfs_zone_finish_endio() does additional checks if the block group can be
finished or not.
But in the context of btrfs_zone_finish_endio_workfn() only the actual
call to do_zone_finish() is of interest, as the skipping condition when
there is still room to allocate from the block group cannot be checked.
Directly call do_zone_finish() on the block group.
Reviewed-by: Damien Le Moal <dlemoal@kernel.org>
Signed-off-by: Johannes Thumshirn <johannes.thumshirn@wdc.com>
Signed-off-by: David Sterba <dsterba@suse.com>
There's one only one caller of unaccount_log_buffer() and both this
function and the caller are short, so move its code into the caller.
Reviewed-by: Boris Burkov <boris@bur.io>
Reviewed-by: Qu Wenruo <wqu@suse.com>
Signed-off-by: Filipe Manana <fdmanana@suse.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Instead of extracting again the disk_bytenr and disk_num_bytes values from
the file extent item to pass to btrfs_qgroup_trace_extent(), use the key
local variable 'ins' which already has those values, reducing the size of
the source code.
Reviewed-by: Boris Burkov <boris@bur.io>
Reviewed-by: Qu Wenruo <wqu@suse.com>
Signed-off-by: Filipe Manana <fdmanana@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Instead of having an if statement to check for regular and prealloc
extents first, process them in a block, and then following with an else
statement to check for an inline extent, check for an inline extent first,
process it and jump to the 'update_inode' label, allowing us to avoid
having the code for processing regular and prealloc extents inside a
block, reducing the high indentation level by one and making the code
easier to read and avoid line splittings too.
Reviewed-by: Boris Burkov <boris@bur.io>
Reviewed-by: Qu Wenruo <wqu@suse.com>
Signed-off-by: Filipe Manana <fdmanana@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
At replay_one_extent(), we can jump to the code that updates the file
extent range and updates the inode when processing a file extent item that
represents a hole and we don't have the NO_HOLES feature enabled. This
helps reduce the high indentation level by one in replay_one_extent() and
avoid splitting some lines to make the code easier to read.
Reviewed-by: Boris Burkov <boris@bur.io>
Reviewed-by: Qu Wenruo <wqu@suse.com>
Signed-off-by: Filipe Manana <fdmanana@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
In the replay_one_buffer() log tree walk callback we return errors to the
log tree walk caller and then the caller aborts the transaction, if we
have one, or turns the fs into error state if we don't have one. While
this reduces code it makes it harder to figure out where exactly an error
came from. So add the transaction aborts after every failure inside the
replay_one_buffer() callback and the functions it calls, making it as
fine grained as possible, so that it helps figuring out why failures
happen.
Reviewed-by: Boris Burkov <boris@bur.io>
Reviewed-by: Qu Wenruo <wqu@suse.com>
Signed-off-by: Filipe Manana <fdmanana@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
