- 07 May, 2024 40 commits
-
-
Josef Bacik authored
Now that we've pushed the lock_extent() into cow_file_range() we can push the extent locking into cow_file_range_inline() and move the lock_extent in cow_file_range() to after we call cow_file_range_inline(). Reviewed-by:
Goldwyn Rodrigues <rgoldwyn@suse.com> Signed-off-by:
Josef Bacik <josef@toxicpanda.com> Signed-off-by:
David Sterba <dsterba@suse.com>
-
Josef Bacik authored
Now that cow_file_range is the only function that is called with the range locked, push this call into cow_file_range so we can further narrow the scope. Reviewed-by:
-
Josef Bacik authored
This is used by zoned but also as the fallback for uncompressed extents when we fail to compress the ranges. Push the extent lock into run_dealloc_cow(), and adjust the compression case to take the extent lock after calling run_delalloc_cow(). Reviewed-by:
-
Josef Bacik authored
Since we immediately unlock the extent range when we enter run_delalloc_compressed() simply move the lock_extent() down to cover cow_file_range() and then remove the unlock_extent() from run_delalloc_compressed. Reviewed-by:
-
Josef Bacik authored
run_delalloc_nocow is a little special because we use the file extents to see if we can nocow a range. We don't actually need the protection of the extent lock to look at the file extents at this point however. We are currently holding the page lock for this range, so we are protected from anybody who would simultaneously be modifying the file extent items for this range. * mmap() - we're holding the page lock. * buffered writes - we're holding the page lock. * direct writes - we're holding the page lock and direct IO has to flush page cache before it's able to continue. * fallocate() - all callers flush the range and wait on ordered extents while holding the inode lock and the mmap lock, so we are again saved by the page lock. We want to use the extent lock to protect 1) The mapping tree for the given range. 2) The ordered extents for the given range. 3) The io_tree for the given range. Push the extent lock down to cover these operations. In the fallback_to_cow() case we simply lock before doing anything and rely on the cow_file_range() helper to handle it's range properly. Reviewed-by:
-
Josef Bacik authored
We have the following pattern while (1) { if (cur_offset > end) break; } Which is just while (cur_offset <= end) { ... } so adjust the code to be more clear. Reviewed-by: -
Josef Bacik authored
run_delalloc_nocow is a bit special as it walks through the file extents for the inode and determines what it can nocow and what it can't. This is the more complicated area for extent locking, so start with this function. Reviewed-by:
-
Josef Bacik authored
We want to limit the scope of the extent lock to be around operations that can change in flight. Currently we hold the extent lock through the entire writepage operation, which isn't really necessary. We want to protect to make sure nobody has updated DELALLOC. In find_lock_delalloc_range we must lock the range in order to validate the contents of our io_tree. However once we've done that we're safe to unlock the range and continue, as we have the page lock already held for the range. We are protected from all operations at this point. * mmap() - we're holding the page lock, thus are protected. * buffered writes - again, we're protected because we take the page lock for the first and last page in our range for buffered writes so we won't create new delalloc ranges in this area. * direct IO - we invalidate pagecache before attempting to write a new area, which requires the page lock, so again are protected once we're holding the page lock on this range. Additionally this behavior actually already exists for compressed, we unlock the range as soon as we start to process the async extents, and re-lock it during compression. So this is completely safe, and makes the locking more consistent. Make this simple by just pushing the extent lock into btrfs_run_delalloc_range. From there followup patches will push the lock further down into its users. Reviewed-by:
-
Josef Bacik authored
We currently don't lock the extent when we're doing a cow_file_range_inline() for a compressed extent. This isn't a problem necessarily, but it's inconsistent with the rest of our usage of cow_file_range_inline(). This also leads to some extra weird logic around whether the extent is locked or not. Fix this to lock the extent before calling cow_file_range_inline() in compression to make it consistent with the rest of the inline users. In future patches this will be pushed down into the cow_file_range_inline() helper, so we're fine with the quick and dirty locking here. This patch exists to make the behavior change obvious. Signed-off-by:
-
Josef Bacik authored
We duplicate the extent cleanup for cow_file_range_inline() in the cow and compressed case. The encoded case doesn't need to do cleanup the same way, so rename cow_file_range_inline to __cow_file_range_inline and then make cow_file_range_inline handle the extent cleanup appropriately, and update the callers. Reviewed-by:
Johannes Thumshirn <johannes.thumshirn@wdc.com> Signed-off-by:
-
Josef Bacik authored
Since 4750af3b ("btrfs: prevent extent_clear_unlock_delalloc() to unlock page not locked by __process_pages_contig()") we have been unlocking the locked page manually instead of via extent_clear_unlock_delalloc() because of subpage blocksize support. However we actually disable inline extent creation for subpage blocksize support, so this behavior isn't necessary. Remove this code and comment, if at some point the subpage blocksize code grows support for inline extents this can be re-evaluated. Reviewed-by:
-
Josef Bacik authored
Currently we have a lot of duplicated checks of if (start == 0 && fs_info->sectorsize == PAGE_SIZE) cow_file_range_inline(); Instead of duplicating this check everywhere, consolidate all of the inline extent logic into a helper which documents all of the checks and then use that helper inside of cow_file_range_inline(). With this we can clean up all of the calls to either unconditionally call cow_file_range_inline(), or at least reduce the checks we're doing before we call cow_file_range_inline(); Reviewed-by:
-
Josef Bacik authored
In the cow path we will clone the reloc csums for relocated data extents, and if there's an error we already have an ordered extent and rely on the ordered extent finishing to clean everything up. There's a problem however, we don't mark the ordered extent with an error, we pretend like everything was just fine. If we were at the end of our range we won't actually bubble up this error anywhere, and we could end up inserting an extent that doesn't have csums where it should have them. Fix this by adding a helper to mark the ordered extent with an error, and then use this when we fail to lookup the csums in btrfs_reloc_clone_csums. Use this helper in the other place where we use the same pattern while we're here. This will prevent us from erroneously inserting the extent that doesn't have the required checksums. Reviewed-by:
-
Qu Wenruo authored
The function create_io_em() is called before we submit an IO, to update the in-memory extent map for the involved range. This patch changes the following aspects: - Does not allow BTRFS_ORDERED_NOCOW type For real NOCOW (excluding NOCOW writes into preallocated ranges) writes, we never call create_io_em(), as we does not need to update the extent map at all. So remove the sanity check allowing BTRFS_ORDERED_NOCOW type. - Add extra sanity checks * PREALLOC - @block_len == len For uncompressed writes. * REGULAR - @block_len == @orig_block_len == @ram_bytes == @len We're creating a new uncompressed extent, and referring all of it. - @orig_start == @start We haven no offset inside the extent. * COMPRESSED - valid @compress_type - @len <= @ram_bytes This is to co-operate with encoded writes, which can cause a new file extent referring only part of a uncompressed extent. Reviewed-by:Filipe Manana <fdmanana@suse.com> Signed-off-by:
Qu Wenruo <wqu@suse.com> Reviewed-by:
-
Qu Wenruo authored
With the tree-checker ensuring all inline file extents starts at file offset 0 and has a length no larger than sectorsize, we can simplify the calculation to assigned those fixes values directly. Reviewed-by:
-
Qu Wenruo authored
The extent_map structure is very critical to btrfs, as it is involved for both read and write paths. Unfortunately the structure is not properly explained, making it pretty hard to understand nor to do further improvement. This patch adds extra comments explaining the major members based on my code reading. Hopefully we can find more members to cleanup in the future. Reviewed-by:
-
Naohiro Aota authored
calcu_metadata_size() has a "reserve" argument, but the only caller always set it to "1". The other usage (reserve = 0) is dropped by a commit 0647bf56 ("Btrfs: improve forever loop when doing balance relocation"), which is more than 10 years ago. Drop the argument and simplify the code. Reviewed-by:
Naohiro Aota <naohiro.aota@wdc.com> Reviewed-by:
-
Anand Jain authored
There's another return variable wret that is only passed to ret on error, we can simply use ret. Signed-off-by:
Anand Jain <anand.jain@oracle.com> Reviewed-by:
-
Anand Jain authored
First, drop err instead reuse ret, choose to return the error instead of goto fail and then return the same error. Do not initialize the ret until where it has to be initialized. Slight logic change in handling the btrfs_search_slot() and btrfs_next_leaf() return value. Signed-off-by:
-
Anand Jain authored
Rename ret to ret2 compile and then err to ret. Also, new ret2 is found to be localized within the 'if (trans)' statement, so move its declaration there. Signed-off-by:
-
Anand Jain authored
In quick_update_accounting() err is used as 2nd return value, which could be achieved just with ret. Signed-off-by:
-
Anand Jain authored
Coding style fixes the function relocate_tree_blocks(). After the fix, ret is the return value variable. Signed-off-by:
-
Anand Jain authored
Code style fix in the function build_backref_tree(). Drop the ret initialization 0, as we don't need it. Signed-off-by:
-
Anand Jain authored
Rename the function's local return variables err and werr to ret. Also, align the variable declarations with the other declarations in the function for better function space alignment. Signed-off-by:
-
Anand Jain authored
Rename the function's local variable werr and err to ret. Reviewed-by:
-
Anand Jain authored
In the function btrfs_write_marked_extents() and in __btrfs_wait_marked_extents() return the actual error if when filemap_fdata<write|wait>_range() fails. Suggested-by:
-
David Sterba authored
There are open coded tests of BTRFS_FS_STATE_DUMMY_FS_INFO and we have a wrapper for that that's a compile-time constant when self-tests are not built in. As this is only for development we can save some bytes and conditions on release configs by using the helper in the remaining cases. Reviewed-by:
-
Filipe Manana authored
There's no need to initialize the delayed inodes xarray with a GFP_ATOMIC flag because that actually does nothing on the xarray operations. That was needed for radix trees, but for xarrays the allocation flags are passed as the last argument to xa_store() (which we are using correctly). So initialize the delayed inodes xarray with a simple xa_init(). Reviewed-by:
-
Filipe Manana authored
Currently try_release_extent_mapping() as an int return type, but we use it as a boolean. Its only caller, the release folio callback, also returns a boolean which corresponds to try_release_extent_mapping()'s return value. So change its return value type to bool as well as its helper try_release_extent_state(). Reviewed-by:
-
Filipe Manana authored
At try_release_extent_mapping(), called during the release folio callback (btrfs_release_folio() callchain), we don't release any extent maps in the range if the GFP flags don't allow blocking. This behaviour is exaggerated because: 1) Both searching for extent maps and removing them are not blocking operations. The only thing that it is the cond_resched() call at the end of the loop that searches for and removes extent maps; 2) We currently only operate on a single page, so for the case where block size matches the page size, we can only have one extent map, and for the case where the block size is smaller than the page size, we can have at most 16 extent maps. So it's very unlikely the cond_resched() call will ever block even in the block size smaller than page size scenario. So instead of not removing any extent maps at all in case the GFP glags don't allow blocking, keep removing extent maps while we don't need to reschedule. This makes it safe for the subpage case and for a future where we can process folios with a size larger than a page. Reviewed-by:
-
Filipe Manana authored
Currently we don't attempt to release extent maps if the inode has an i_size that is not greater than 16M. This condition was added way back in 2008 by commit 70dec807 ("Btrfs: extent_io and extent_state optimizations"), without any explanation about it. A quick chat with Chris on slack revealed that the goal was probably to release the extent maps for small files only when closing the inode. This however can be harmful in case we have tons of such files being kept open for very long periods of time, since we will consume more and more pages for extent maps. So remove the condition. Reviewed-by:
-
Filipe Manana authored
Nowadays we have the btrfs_get_fs_generation() to get the current generation of the filesystem, so there's no need anymore to lock the transaction spinlock to read it. Reviewed-by:
-
Filipe Manana authored
Rename the following variables: 1) "btrfs_inode" to "inode", because it's shorter to type and clear, and we don't have a VFS inode here as well, so there's no confusion; 2) "tree" to "io_tree", to be clear which tree we are dealing with, since we use 2 different trees in the function; 3) "map" to "extent_tree" since "map" gives the idea we are dealing with an extent map for example, but we are dealing with the inode's extent tree (the tree which stores extent maps). These also make the next patches simpler. Reviewed-by:
-
Filipe Manana authored
Add some tracepoints for the extent map shrinker to help debug and analyse main events. These have proved useful during development of the shrinker. Reviewed-by:
-
Filipe Manana authored
Nowadays we have a lock used to synchronize mmap writes with reflink and fsync operations (struct btrfs_inode::i_mmap_lock), so update the comment for btrfs_set_inode_full_sync() to mention that it can also be called while holding that mmap lock. Besides being a valid alternative to the inode's VFS lock, we already have the extent map shrinker using that mmap lock instead. Reviewed-by:
-
Filipe Manana authored
Extent maps are used either to represent existing file extent items, or to represent new extents that are going to be written and the respective file extent items are created when the ordered extent completes. We currently don't have any limit for how many extent maps we can have, neither per inode nor globally. Most of the time this not too noticeable because extent maps are removed in the following situations: 1) When evicting an inode; 2) When releasing folios (pages) through the btrfs_release_folio() address space operation callback. However we won't release extent maps in the folio range if the folio is either dirty or under writeback or if the inode's i_size is less than or equals to 16M (see try_release_extent_mapping(). This 16M i_size constraint was added back in 2008 with commit 70dec807 ("Btrfs: extent_io and extent_state optimizations"), but there's no explanation about why we have it or why the 16M value. This means that for buffered IO we can reach an OOM situation due to too many extent maps if either of the following happens: 1) There's a set of tasks constantly doing IO on many files with a size not larger than 16M, specially if they keep the files open for very long periods, therefore preventing inode eviction. This requires a really high number of such files, and having many non mergeable extent maps (due to random 4K writes for example) and a machine with very little memory; 2) There's a set tasks constantly doing random write IO (therefore creating many non mergeable extent maps) on files and keeping them open for long periods of time, so inode eviction doesn't happen and there's always a lot of dirty pages or pages under writeback, preventing btrfs_release_folio() from releasing the respective extent maps. This second case was actually reported in the thread pointed by the Link tag below, and it requires a very large file under heavy IO and a machine with very little amount of RAM, which is probably hard to happen in practice in a real world use case. However when using direct IO this is not so hard to happen, because the page cache is not used, and therefore btrfs_release_folio() is never called. Which means extent maps are dropped only when evicting the inode, and that means that if we have tasks that keep a file descriptor open and keep doing IO on a very large file (or files), we can exhaust memory due to an unbounded amount of extent maps. This is especially easy to happen if we have a huge file with millions of small extents and their extent maps are not mergeable (non contiguous offsets and disk locations). This was reported in that thread with the following fio test: $ cat test.sh #!/bin/bash DEV=/dev/sdj MNT=/mnt/sdj MOUNT_OPTIONS="-o ssd" MKFS_OPTIONS="" cat <<EOF > /tmp/fio-job.ini [global] name=fio-rand-write filename=$MNT/fio-rand-write rw=randwrite bs=4K direct=1 numjobs=16 fallocate=none time_based runtime=90000 [file1] size=300G ioengine=libaio iodepth=16 EOF umount $MNT &> /dev/null mkfs.btrfs -f $MKFS_OPTIONS $DEV mount $MOUNT_OPTIONS $DEV $MNT fio /tmp/fio-job.ini umount $MNT Monitoring the btrfs_extent_map slab while running the test with: $ watch -d -n 1 'cat /sys/kernel/slab/btrfs_extent_map/objects \ /sys/kernel/slab/btrfs_extent_map/total_objects' Shows the number of active and total extent maps skyrocketing to tens of millions, and on systems with a short amount of memory it's easy and quick to get into an OOM situation, as reported in that thread. So to avoid this issue add a shrinker that will remove extents maps, as long as they are not pinned, and takes proper care with any concurrent fsync to avoid missing extents (setting the full sync flag while in the middle of a fast fsync). This shrinker is triggered through the callbacks nr_cached_objects and free_cached_objects of struct super_operations. The shrinker will iterate over all roots and over all inodes of each root, and keeps track of the last scanned root and inode, so that the next time it runs, it starts from that root and from the next inode. This is similar to what xfs does for its inode reclaim (implements those callbacks, and cycles through inodes by starting from where it ended last time). Reviewed-by:
-
Filipe Manana authored
Add a per cpu counter that tracks the total number of extent maps that are in extent trees of inodes that belong to fs trees. This is going to be used in an upcoming change that adds a shrinker for extent maps. Only extent maps for fs trees are considered, because for special trees such as the data relocation tree we don't want to evict their extent maps which are critical for the relocation to work, and since those are limited, it's not a concern to have them in memory during the relocation of a block group. Another case are extent maps for free space cache inodes, which must always remain in memory, but those are limited (there's only one per free space cache inode, which means one per block group). Reviewed-by:
-
Filipe Manana authored
Extent maps are always associated to an inode's extent map tree, so there's no need to pass the extent map tree explicitly to try_merge_map(). In order to facilitate an upcoming change that adds a shrinker for extent maps, change try_merge_map() to receive the inode instead of its extent map tree. Reviewed-by:
-
Filipe Manana authored
Extent maps are always associated to an inode's extent map tree, so there's no need to pass the extent map tree explicitly to setup_extent_mapping(). In order to facilitate an upcoming change that adds a shrinker for extent maps, change setup_extent_mapping() to receive the inode instead of its extent map tree. Reviewed-by:
-
Filipe Manana authored
Extent maps are always associated to an inode's extent map tree, so there's no need to pass the extent map tree explicitly to replace_extent_mapping(). In order to facilitate an upcoming change that adds a shrinker for extent maps, change replace_extent_mapping() to receive the inode instead of its extent map tree. Reviewed-by:
-
