kernel.git/fs/bcachefs/six.c, branch linux-6.7.yHosts the 0x221E linux distro kernel.
https://git.0xinfinity.dev/distro/kernel.git/atom?h=linux-6.7.y2023-11-15T04:44:44Zbcachefs: six locks: Fix lost wakeup2023-11-15T04:44:44ZKent Overstreet2023-11-14T23:52:22Zurn:sha1:61b85cb0d773115d9a4b20c3e67286844cf73f34
In percpu reader mode, trylock() for read had a lost wakeup: on failure
to get the lock, we may have caused a writer to fail to get the lock,
because we temporarily elevated the reader count.
We need to check for waiters after decrementing the read count - not
before.
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
six locks: Lock contended tracepoints2023-10-31T02:26:45ZKent Overstreet2023-02-05T21:18:59Zurn:sha1:20e425d301d673dbd5df0c9d4b186c70b43813bb
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
bcachefs: Fix W=12 build errors2023-10-22T21:10:13ZKent Overstreet2023-09-12T22:41:22Zurn:sha1:96dea3d599dbc31f59eb786af2ac5079122beb88
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
bcachefs: six locks: Guard against wakee exiting in __six_lock_wakeup()2023-10-22T21:10:10ZKent Overstreet2023-08-12T21:10:42Zurn:sha1:f854ce4d0a3f7281b5b99b28dd028abe21c1b0c8
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
bcachefs: six locks: Fix missing barrier on wait->lock_acquired2023-10-22T21:10:10ZKent Overstreet2023-08-12T19:05:06Zurn:sha1:c294ea50da4b1a0ee84253f46391aa87a6efe91c
Six locks do lock handoff via the wakeup path: the thread doing the
wakeup also takes the lock on behalf of the waiter, which means the
waiter only has to look at its waitlist entry, and doesn't have to touch
the lock cacheline while another thread is using it.
Linus noticed that this needs a real barrier, which this patch fixes.
Also add a comment for the should_sleep_fn() error path.
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Boqun Feng <boqun.feng@gmail.com>
Cc: linux-bcachefs@vger.kernel.org
Cc: linux-kernel@vger.kernel.org
six locks: Disable percpu read lock mode in userspace2023-10-22T21:10:02ZKent Overstreet2023-06-10T14:57:23Zurn:sha1:aab5e0972a32790c0dbfac64929529820114c674
When running in userspace, we currently don't have a real percpu
implementation available - at least in bcachefs-tools, which is where
this code is currently used in userspace.
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
six locks: Use atomic_try_cmpxchg_acquire()2023-10-22T21:10:02ZKent Overstreet2023-05-25T18:35:06Zurn:sha1:2d9200cfe004b8208a04ea15b3967f9a14135be6
This switches to a newer cmpxchg variant which updates @old for us on
failure, simplifying the cmpxchg loops a bit and supposedly generating
better code.
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
six locks: Fix an unitialized var2023-10-22T21:10:02ZKent Overstreet2023-05-25T22:10:04Zurn:sha1:c4687a4a7536c3b2139faa63e66afd1d3da5bf15
In the conversion to atomic_t, six_lock_slowpath() ended up calling
six_lock_wakeup() in the failure path with a state variable that was
never initialized - whoops.
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
six locks: Delete redundant comment2023-10-22T21:10:02ZKent Overstreet2023-05-23T04:21:22Zurn:sha1:96e53e909d0433f73831315b106f16895a74b843
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
six locks: Tiny bit more tidying2023-10-22T21:10:02ZKent Overstreet2023-05-22T16:11:13Zurn:sha1:2ab62310fd1c723bd8ab8e8242e31fa494c9681f
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>