Commits · 690f7cdf732a032388e26c5daed6ad46c9e48256 · Kirill Smelkov / linux

09 May, 2024 4 commits
- bcachefs: bch2_dev_get_ioref2(); btree_io.c · 690f7cdf
  Kent Overstreet authored May 03, 2024
```
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
```
  690f7cdf
- bcachefs: bch2_dev_get_ioref2(); backpointers.c · 466298e2
  Kent Overstreet authored May 03, 2024
```
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
```
  466298e2
- bcachefs: bch2_dev_get_ioref2(); alloc_background.c · 0e57996c
  Kent Overstreet authored May 03, 2024
```
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
```
  0e57996c
- bcachefs: for_each_bset() declares loop iter · b6fb4269
  Kent Overstreet authored May 08, 2024
```
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
```
  b6fb4269
08 May, 2024 36 commits

bcachefs: Move BCACHEFS_STATFS_MAGIC value to UAPI magic.h · e2f48c48

Petr Vorel authored May 07, 2024

Move BCACHEFS_STATFS_MAGIC value to UAPI <linux/magic.h> under
BCACHEFS_SUPER_MAGIC definition (use common approach for name) and reuse the
definition in bcachefs_format.h BCACHEFS_STATFS_MAGIC.

There are other bcachefs magic definitions: BCACHE_MAGIC, BCHFS_MAGIC,
which use UUID_INIT() and are used only in libbcachefs. Therefore move
only BCACHEFS_STATFS_MAGIC value, which can be used outside of
libbcachefs for f_type field in struct statfs in statfs() or fstatfs().
Suggested-by: Su Yue <glass.su@suse.com>
Signed-off-by: Petr Vorel <pvorel@suse.cz>
Acked-by: Brian Foster <bfoster@redhat.com>
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>

e2f48c48

bcachefs: Improve sysfs internal/btree_cache · e11ecc61
Kent Overstreet authored May 05, 2024
```
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
```
e11ecc61

bcachefs: Allocator prefers not to expand mi.btree_allocated bitmap · c6705091

Kent Overstreet authored Apr 20, 2024

We now have a small bitmap in the member info section of the superblock
for "regions that have btree nodes", so that if we ever have to scan for
btree nodes in repair we don't have to scan the whole device(s).

This tweaks the allocator to prefer allocating from regions that are
already marked in this bitmap.
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>

c6705091

bcachefs: Better bucket alloc tracepoints · 40574946

Kent Overstreet authored May 04, 2024

Tracepoints are garbage, and perf trace even cuts off some of our
fields.

Much nicer to just trace a string, and then we can build nicely
formatted output with printbufs.
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>

40574946

bcachefs: Move nocow unlock to bch2_write_endio() · 62025697

Kent Overstreet authored May 01, 2024

This fixes a lifetime issue; bch2_nocow_write_unlock() uses
PTR_BUCKET_POS(), which needs the device - but we drop our ref to the
device in bch2_write_endio().
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>

62025697

bcachefs: kill bch2_dev_bkey_exists() in journal_ptrs_to_text() · b6d29b58
Kent Overstreet authored May 01, 2024
```
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
```
b6d29b58
bcachefs: kill bch2_dev_bkey_exists() in discard_one_bucket_fast() · f4301b63
Kent Overstreet authored May 01, 2024
```
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
```
f4301b63
bcachefs: kill bch2_dev_bkey_exists() in check_alloc_info() · bc3204c8
Kent Overstreet authored May 01, 2024
```
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
```
bc3204c8

bcachefs: bch2_dev_have_ref() · d8585a79

Kent Overstreet authored May 01, 2024

bch2_dev_bkey_exists() is going away; bch2_dev_have_ref() documents that
we're looking up a device without checking if it's present because we
have a reference to it already.
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>

d8585a79

bcachefs: kill bch2_dev_bkey_exists() in data_update_init() · 222eacab
Kent Overstreet authored May 01, 2024
```
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
```
222eacab
bcachefs: kill bch2_dev_bkey_exists() in bkey_pick_read_device() · a9422fd4
Kent Overstreet authored Apr 30, 2024
```
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
```
a9422fd4
bcachefs: pass bch_dev to read_from_stale_dirty_pointer() · db39a35d
Kent Overstreet authored May 01, 2024
```
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
```
db39a35d
bcachefs: bch2_dev_bucket_exists() uses bch2_dev_rcu() · 78e9b548
Kent Overstreet authored May 01, 2024
```
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
```
78e9b548
bcachefs: kill bch2_dev_bkey_exists() in btree_gc.c · ad897d24
Kent Overstreet authored Apr 30, 2024
```
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
```
ad897d24
bcachefs: bch2_extent_normalize() -> bch2_dev_rcu() · 9cadb4ea
Kent Overstreet authored Apr 30, 2024
```
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
```
9cadb4ea
bcachefs: bch2_bkey_has_target() -> bch2_dev_rcu() · 8e3cc200
Kent Overstreet authored Apr 30, 2024
```
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
```
8e3cc200
bcachefs: extent_ptr_invalid() -> bch2_dev_rcu() · 8feecbed
Kent Overstreet authored Apr 30, 2024
```
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
```
8feecbed
bcachefs: ptr_stale() -> dev_ptr_stale() · 3858aa42
Kent Overstreet authored Apr 30, 2024
```
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
```
3858aa42
bcachefs: extent_ptr_durability() -> bch2_dev_rcu() · 302c980a
Kent Overstreet authored Apr 30, 2024
```
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
```
302c980a
bcachefs: bch2_extent_merge() -> bch2_dev_rcu() · 3793b3f9
Kent Overstreet authored Apr 30, 2024
```
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
```
3793b3f9
bcachefs: ec_validate_checksums() -> bch2_dev_tryget() · c387d844
Kent Overstreet authored Apr 30, 2024
```
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
```
c387d844

bcachefs: ob_dev() · 8783856a

Kent Overstreet authored Apr 30, 2024

Wrapper around bch2_dev_have_ref() for open_buckets; we do guarantee
that the device an open_bucket points to exists.
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>

8783856a

bcachefs: move replica_set from bch_dev to bch_fs · dbd04080

Kent Overstreet authored Apr 30, 2024

This is needed for the next patch - the write submit path has to be able
to allocate a replica bio even when we weren't able to get a ref on the
device.
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>

dbd04080

bcachefs: Kill bch2_dev_bkey_exists() in backpointer code · 633cf069
Kent Overstreet authored Apr 30, 2024
```
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
```
633cf069
bcachefs: PTR_BUCKET_POS() now takes bch_dev · 1f2f92ec
Kent Overstreet authored Apr 30, 2024
```
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
```
1f2f92ec

bcachefs: bch2_dev_iterate() · fa6cce09

Kent Overstreet authored Apr 30, 2024

New helper for getting refs to devices as we iterate.
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>

fa6cce09

bcachefs: bch2_evacuate_bucket() -> bch2_dev_tryget() · cb4d340a
Kent Overstreet authored Apr 30, 2024
```
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
```
cb4d340a
bcachefs: bch2_bucket_ref_update() now takes bch_dev · 07d7c4da
Kent Overstreet authored Apr 30, 2024
```
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
```
07d7c4da
bcachefs: bch2_trigger_alloc() -> bch2_dev_tryget() · a7f1c26f
Kent Overstreet authored Apr 30, 2024
```
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
```
a7f1c26f
bcachefs: bch2_check_alloc_key() -> bch2_dev_tryget_noerror() · 9b3059a1
Kent Overstreet authored Apr 30, 2024
```
More elimination of bch2_dev_bkey_exists() usage.
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
```
9b3059a1
bcachefs: Convert to bch2_dev_tryget_noerror() · 4cd91e2f
Kent Overstreet authored May 03, 2024
```
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
```
4cd91e2f

bcachefs: bch2_dev_tryget() · b07eb825

Kent Overstreet authored Apr 30, 2024

Most uses of bch2_dev_bkey_exists() are going away, where we assume that
because a key references a device the device most exist - instead, we'll
be explicitly checking if the device exists and getting a reference to
it.

This adds the new helpers.
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>

b07eb825

bcachefs: bch2_have_enough_devs() checks for nonexistent device · 6349b07c
Kent Overstreet authored Apr 30, 2024
```
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
```
6349b07c
bcachefs: journal_replay_entry_early() checks for nonexistent device · adf81796
Kent Overstreet authored Apr 30, 2024
```
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
```
adf81796
bcachefs: bch2_dev_btree_bitmap_marked() -> bch2_dev_rcu() · 13a16dab
Kent Overstreet authored Apr 30, 2024
```
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
```
13a16dab

bcachefs: Pass device to bch2_bucket_do_index() · 267039d0

Kent Overstreet authored Apr 30, 2024

Eliminating bch2_dev_bkey_exists() uses and replacing them with proper
checks; this one was unnecessary since the caller already has it.
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>

267039d0