Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>

prep work for persistent alloc info
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>

Will be used in the future for inode updates, which will be very helpful
for multithreaded workloads that have to update the inode with every
extent update (appends, or updates that change i_sectors)

Also will be used eventually for fully persistent alloc info

However - we still need a mechanism for reserving space in the journal
prior to getting a journal reservation, so it's not technically safe to
make use of this just yet, we could deadlock with the journal full
(although not likely to be an issue in practice)
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>

caller should already be guarding against rw, and checking here breaks
when caller needs to finish updates for going RO
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>

Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>

this lets us get rid of a lot of extra switch statements - in a lot of
places we dispatch on the btree node type, and then the key type, so
this is a nice cleanup across a lot of code.

Also improve the on disk format versioning stuff.
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>

previously, if the code traversed to the next btree node, that could
return an error (due to lock restarts) - which was not being checked
for.

fix is to rework it so it never iterates past the current leaf node, and
pops an assertion if it ever sees an error.
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>

Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>

Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>

Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>

Fixes an inconsistency at the end of gc
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>

gc now verifies the contents of the stripes radix tree, important for
persistent alloc info
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>

Hit an assertion, probably spurious, indicating an iterator was unlocked
when it shouldn't have been (spurious because it wasn't locked at all
when the caller called btree_insert_at()).

Add a flag, BTREE_ITER_NOUNLOCK, and tighten up the assertions
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>

Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>

This reverts commit 36f389604294dfc953e6f5624ceb683818d32f28.
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>

Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>

This means we can now use gc to verify the allocation information -
important for testing persistant alloc info
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>

Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>

Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>

Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>

Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>

Fixes a deadlock between the allocator thread, when it first starts up,
and journal replay
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>

Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>

Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>

Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>

Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>

Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>

more prep work for erasure coding
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>

Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>

Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>

Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>

Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>

Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>

Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>

Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>

This way the last mount time is actually meaningful instead of just being
various times from 1970 (which happens with the monotonic clock).

Also, roundup_pow_of_two() is undefined when passed in 0, so check before
calling it.
Signed-off-by: Tim Schlueter <schlueter.tim@linux.com>
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>

Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>

It's now possible to create and use a filesystem on a 512k device with
4k buckets (though at that size we still waste almost half to internal
reserves)
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>

this means we don't have to reserve space for them when calculating
filesystem capacity
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>

Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>