fs/btrfs/inode.c

386  * Cleanup all submitted ordered extents in specified range to handle errors
534 	 * We align size to sectorsize for inline extents just for simplicity
567 	/* Inline extents must start at offset 0. */
571 	/* Inline extents are limited to sectorsize. */
587 	/* Inline extents must be the entirety of the file. */
739 	struct list_head extents;
768 	list_add_tail(&async_extent->list, &cow->extents);
1242  * allocate extents on disk for the range, and create ordered data structs
1243  * in ram to track those extents.
1252  * (In reality inline extents are limited to a single page, so locked_folio is
1262  * while-loop, the ordered extents created in previous iterations are cleaned up.
1327 	 * Relocation relies on the relocated extents to have exactly the same
1328 	 * size as the original extents. Normally writeback for relocation data
1329 	 * extents follows a NOCOW path because relocation preallocates the
1330 	 * extents. However, due to an operation such as scrub turning a block
1333 	 * not be split into smaller extents, otherwise relocation breaks and
1371 				 * and exit the loop to unlock the processed extents.
1478 	 * For the range (1). We have already instantiated the ordered extents
1479 	 * for this region, thus we need to cleanup those ordered extents.
1481 	 * are also handled by the ordered extents cleanup.
1544  * async extents created by compress_file_range and send them down to the disk.
1574 	while (!list_empty(&async_chunk->extents)) {
1575 		async_extent = list_first_entry(&async_chunk->extents,
1625 		INIT_LIST_HEAD(&async_chunk[i].extents);
1741 	 * reason. Space caches and relocated data extents always get a prealloc
1768 	 * Don't try to create inline extents, as a mix of inline extent that
1848 	/* Compressed/encrypted/encoded extents must be COWed. */
2012 	 * On error, we need to cleanup the ordered extents we created.
2024  * of the extents that exist in the file, and COWs the file as required.
2111 		 * more extents for this inode
2243 	 *    Need to clear the dirty flags and finish the ordered extents.
2267 	 * So for all above cases, if @cow_start is set, cleanup ordered extents
2392  * Handle merged delayed allocation extents so we can keep track of new extents
2393  * that are just merged onto old extents, such as when we are doing sequential
2423 	 * We have to add up either side to figure out how many extents were
2425 	 * extents we accounted for is <= the amount we need for the new range
2431 	 * need 2 outstanding extents, on one side we have 1 and the other side
2436 	 * Each range on their own accounts for 2 extents, but merged together
2437 	 * they are only 3 extents worth of accounting, so we need to drop in
2708 		 * There can't be any extents following eof in this case so just
2866  * to fix it up.  The async helper will wait for ordered extents, set
3794 		 * - extents,
3959 	 * idea about which extents were modified before we were evicted from
4811 	 * We do not need to lock extents nor wait for OE, as it's already
5085  * This function puts in dummy file extents for the area we're creating a hole
5087  * these file extents so that btrfs_get_extent will return a EXTENT_MAP_HOLE for
5258 			 * adjusted disk_i_size down as we removed extents, so
6966  * extents which overlap the given range after the returned extent_map.
7122 		 * Other members are not utilized for inline extents.
7190  * NOTE: This only checks the file extents, caller is responsible to wait for
7191  *	 any ordered extents.
7525 		 * for any ordered extents now. Don't clear EXTENT_DELALLOC_NEW
7585 	 * We have iterated through all ordered extents of the page, the page
7771 	 * write the extents that changed, which is a problem if we need to
7773 	 * all of the extents in the inode to the sync log so we're completely
7776 	 * If no extents were dropped or trimmed we don't need to force the next
7781 	 * extents beyond i_size to drop.
7918 	 * If this is a free space inode do not take the ordered extents lockdep
9568 		 * For inline extents we get everything we need out of the
9699 	 * Compressed extents should always have checksums, so error out if we
9718 	 * extents.
10099 	 * run concurrently while we are mapping the swap extents, and
10101 	 * file is active and moving the extents. Note that this also prevents
10127 	 * Snapshots can create extents which require COW even if NODATACOW is
10129 	 * before walking the extents because we don't want a concurrent
10130 	 * snapshot to run after we've already checked the extents.
10367  * Update the number of bytes used in the VFS' inode. When we replace extents in
10388  * Verify that there are no ordered extents for a given file range.
10398  * extents in the range to complete and finally we have locked the file range in
10489  * use bmap to make a mapping of extents in the file.  They assume
10490  * these extents won't change over the life of the file and they