@@ -193,7 +193,61 @@ features see :doc:`Status<Status>` page.
193
193
not touch the superblock read-only status, preventing removal of
194
194
accumulated deleted snapshots to be cleaned.
195
195
196
- 6.13 - update tree-checker to detect more wrong inline extent references
196
+ 6.13 - tree-checker update for inline extent references
197
+ Update tree-checker to detect more wrong inline extent references.
198
+
199
+ 6.14 - support READ_VERITY_METADATA ioctl
200
+ Add support for FS_IOC_READ_VERITY_METADATA to directly query the
201
+ Merkle tree, descriptor and signature blocks for fs-verity enabled
202
+ files.
203
+
204
+ 6.14 - *experimental * read balancing policies
205
+ Add more read balancing policies, configurable in :file: `/sys/fs/btrfs/FSID/read_policy `
206
+ or as module parameter *read_policy *. Newly added *round-robin *,
207
+ *devid:N * (select a specific mirror number *N *).
208
+
209
+ 6.14 - encoded write integration with io_uring
210
+ The io_uring subsystem understands a command that is directed to
211
+ Btrfs encoded write ioctl.
212
+
213
+ 6.15 - always do buffered IO for files with checksums
214
+ Direct IO may lead to data and their checksums mismatch. Use the direct
215
+ to buffered fallback in case the file has checksums. This has a negative
216
+ performance impact.
217
+
218
+ 6.15 - negative (fast) levels for zstd
219
+ Mount options for *zstd * compression accept negative values *-1..-15 *
220
+ match the levels. They provide faster compression at the cost of worse
221
+ ratio.
222
+
223
+ 6.15 - *(debug builds) * accept 2K block size on x86_64
224
+ For testing the *subpage * block size feature, the size of 2K is accepted
225
+ on x86_64 which has 4K pages.
226
+
227
+ 6.15 - defrag ioctl accepts negative zstd levels
228
+ The defrag ioctl also accepts the negative zstd levels that can be set as
229
+ mount option.
230
+
231
+ 6.17 - track current commit duration in commit_stats
232
+ Add entry to :file: `commit_stats ` to detect commit stalls, for
233
+ debugging or monitoring purposes.
234
+
235
+ 6.17 - *experimental * large folio support
236
+ Large folios abstract contiguous page ranges representing some filesystem
237
+ data or metadata as one structure instead of several ones. This simplifies
238
+ code and has a positive impact on performance. As it touches the core
239
+ data structure it is not enabled by default.
240
+
241
+ 6.17 - restrict writes to mounted devices
242
+ Any btrfs mounted device cannot be opened for writes.
243
+
244
+ 6.17 - defrag ioctl can force no compression
245
+ The defrag ioctl was not able to uncompress a given range, now it's
246
+ possible.
247
+
248
+ 6.17 - send (v2 protocol) uses fallocate for hole punching
249
+ File holes, ranges not representing data, were emulated by a zero
250
+ filled data. This is less efficient than puching holes.
197
251
198
252
5.x
199
253
---
0 commit comments