- 22 Oct, 2023 40 commits
-
-
Kent Overstreet authored
Importantly, we don't want to use bch2_fs_inconsistent_on() for errors that fsck can repair, becuase that will just put us in RO mode and prevent fsck from actually fixing stuff. Probably want to get rid of it in the future. Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
-
Kent Overstreet authored
Right now this is the only way of repairing bucket gens in the future Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
-
Kent Overstreet authored
when iterating over reflink pointers, we use the key we just emitted to set the iterator position - which means we have to be setting the key's inode field as well Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
-
Kent Overstreet authored
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
-
Kent Overstreet authored
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
-
Kent Overstreet authored
The continue statement in bch2_trans_mark_extent() was wrong - by bailing out early, we'd be constructing the wrong replicas list to update. Also, the assertion in update_replicas() was wrong - due to rounding with compressed extents, it is possible for sectors to be 0 sometimes. Also, change extent_to_replicas() in replicas.c to match the replicas list we construct in buckets.c. Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
-
Kent Overstreet authored
Major simplification - gets rid of the need for marking buckets as dirty, instead we write buckets if the in memory mark is different from what's in the btree. Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
-
Kent Overstreet authored
This fixes a bug in the journal replay -> extent_replay_key -> split_compressed path, when we do an update that changes alloc info but the alloc info in the btree isn't up to date yet. Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
-
Kent Overstreet authored
Reflink caused fiemap performance to regress badly - this gets us back to where we were. Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
-
Kent Overstreet authored
This way we aren't doing a full linear scan every time we create a new stripe. Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
-
Kent Overstreet authored
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
-
Kent Overstreet authored
It's only updating timestamps, so this doubly doesn't make sense. fsync will flush the journal, if necessary. Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
-
Kent Overstreet authored
*i is used as an output parameter, but gcc isn't noticing that. Oh well. Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
-
Kent Overstreet authored
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
-
Kent Overstreet authored
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
-
Kent Overstreet authored
Also, move other update path checks to where they actually check all the updates (after triggers have run) Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
-
Kent Overstreet authored
Also - fix alignment in bch2_set_page_dirty() Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
-
Kent Overstreet authored
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
-
Kent Overstreet authored
We have to reinitialize ptrs whenever we do something that changes them. Regression from when the code was converted to be generic across all keys with pointers. Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
-
Kent Overstreet authored
bch2_bkey_normalize() modifies the value, and we were modifying the original value in the src btree node - but, we're called without a write lock held on the src node. Oops... Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
-
Kent Overstreet authored
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
-
Kent Overstreet authored
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
-
Kent Overstreet authored
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
-
Kent Overstreet authored
bch2_btree_node_iter_prev_all() depends on an invariant that wasn't being maintained for extent leaf nodes - specifically, the node iterator may not have advanced past any keys that compare after the key the node iterator points to. Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
-
Kent Overstreet authored
bch2_btree_node_iter_prev_filter() tried to be smart about iterating backwards when skipping over whiteouts/discards - but unfortunately, doing so can leave the node iterator in an inconsistent state; the sane solution is to just always iterate backwards one key at a time. But we compact btree nodes when more than a quarter of the keys are whiteouts/discards, so the optimization wasn't buying us that much anyways. Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
-
Kent Overstreet authored
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
-
Kent Overstreet authored
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
-
Kent Overstreet authored
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
-
Kent Overstreet authored
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
-
Kent Overstreet authored
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
-
Kent Overstreet authored
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
-
Kent Overstreet authored
More prep work for reflink: for extents, we're not looking for an exact mach on pos, rather that the pos is within the range of the key the iterator points to. Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
-
Kent Overstreet authored
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
-
Kent Overstreet authored
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
-
Kent Overstreet authored
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
-
Kent Overstreet authored
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
-
Kent Overstreet authored
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
-
Kent Overstreet authored
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
-
Kent Overstreet authored
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
-
Kent Overstreet authored
With reflink, various code now has to handle both KEY_TYPE_extent or KEY_TYPE_reflink_v - so, convert it to be generic across all keys with pointers. Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
-