closures: use seq_putc() in debug_show()

A single line break should be put into a sequence. Thus use the
corresponding function "seq_putc".

This issue was transformed by using the Coccinelle softw

closures: use seq_putc() in debug_show()

A single line break should be put into a sequence. Thus use the
corresponding function "seq_putc".

This issue was transformed by using the Coccinelle software.

Link: https://lkml.kernel.org/r/[email protected]
Signed-off-by: Markus Elfring <[email protected]>
Cc: Kent Overstreet <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>

show more ...

closures: fix closure_sync + closure debugging

originally, stack closures were only used synchronously, and with the
original implementation of closure_sync() the ref never hit 0; thus,
closure_put_

closures: fix closure_sync + closure debugging

originally, stack closures were only used synchronously, and with the
original implementation of closure_sync() the ref never hit 0; thus,
closure_put_after_sub() assumes that if the ref hits 0 it's on the debug
list, in debug mode.

that's no longer true with the current implementation of closure_sync,
so we need a new magic so closure_debug_destroy() doesn't pop an assert.

Signed-off-by: Kent Overstreet <[email protected]>

show more ...

closures: closure_get_not_zero(), closure_return_sync()

Provide new primitives for solving a lifetime issue with bcachefs
btree_trans objects.

closure_sync_return(): like closure_sync(), wait synch

closures: closure_get_not_zero(), closure_return_sync()

Provide new primitives for solving a lifetime issue with bcachefs
btree_trans objects.

closure_sync_return(): like closure_sync(), wait synchronously for any
outstanding gets. like closure_return, the closure is considered
"finished" and the ref left at 0.

closure_get_not_zero(): get a ref on a closure if it's alive, i.e. the
ref is not zero.

Signed-off-by: Kent Overstreet <[email protected]>

show more ...

closures: Change BUG_ON() to WARN_ON()

If a BUG_ON() can be hit in the wild, it shouldn't be a BUG_ON()

For reference, this has popped up once in the CI, and we'll need more
info to debug it:

0324

closures: Change BUG_ON() to WARN_ON()

If a BUG_ON() can be hit in the wild, it shouldn't be a BUG_ON()

For reference, this has popped up once in the CI, and we'll need more
info to debug it:

03240 ------------[ cut here ]------------
03240 kernel BUG at lib/closure.c:21!
03240 kernel BUG at lib/closure.c:21!
03240 Internal error: Oops - BUG: 00000000f2000800 [#1] SMP
03240 Modules linked in:
03240 CPU: 15 PID: 40534 Comm: kworker/u80:1 Not tainted 6.10.0-rc4-ktest-ga56da69799bd #25570
03240 Hardware name: linux,dummy-virt (DT)
03240 Workqueue: btree_update btree_interior_update_work
03240 pstate: 00001005 (nzcv daif -PAN -UAO -TCO -DIT +SSBS BTYPE=--)
03240 pc : closure_put+0x224/0x2a0
03240 lr : closure_put+0x24/0x2a0
03240 sp : ffff0000d12071c0
03240 x29: ffff0000d12071c0 x28: dfff800000000000 x27: ffff0000d1207360
03240 x26: 0000000000000040 x25: 0000000000000040 x24: 0000000000000040
03240 x23: ffff0000c1f20180 x22: 0000000000000000 x21: ffff0000c1f20168
03240 x20: 0000000040000000 x19: ffff0000c1f20140 x18: 0000000000000001
03240 x17: 0000000000003aa0 x16: 0000000000003ad0 x15: 1fffe0001c326974
03240 x14: 0000000000000a1e x13: 0000000000000000 x12: 1fffe000183e402d
03240 x11: ffff6000183e402d x10: dfff800000000000 x9 : ffff6000183e402e
03240 x8 : 0000000000000001 x7 : 00009fffe7c1bfd3 x6 : ffff0000c1f2016b
03240 x5 : ffff0000c1f20168 x4 : ffff6000183e402e x3 : ffff800081391954
03240 x2 : 0000000000000001 x1 : 0000000000000000 x0 : 00000000a8000000
03240 Call trace:
03240 closure_put+0x224/0x2a0
03240 bch2_check_for_deadlock+0x910/0x1028
03240 bch2_six_check_for_deadlock+0x1c/0x30
03240 six_lock_slowpath.isra.0+0x29c/0xed0
03240 six_lock_ip_waiter+0xa8/0xf8
03240 __bch2_btree_node_lock_write+0x14c/0x298
03240 bch2_trans_lock_write+0x6d4/0xb10
03240 __bch2_trans_commit+0x135c/0x5520
03240 btree_interior_update_work+0x1248/0x1c10
03240 process_scheduled_works+0x53c/0xd90
03240 worker_thread+0x370/0x8c8
03240 kthread+0x258/0x2e8
03240 ret_from_fork+0x10/0x20
03240 Code: aa1303e0 d63f0020 a94363f7 17ffff8c (d4210000)
03240 ---[ end trace 0000000000000000 ]---
03240 Kernel panic - not syncing: Oops - BUG: Fatal exception
03240 SMP: stopping secondary CPUs
03241 SMP: failed to stop secondary CPUs 13,15
03241 Kernel Offset: disabled
03241 CPU features: 0x00,00000003,80000008,4240500b
03241 Memory Limit: none
03241 ---[ end Kernel panic - not syncing: Oops - BUG: Fatal exception ]---
03246 ========= FAILED TIMEOUT copygc_torture_no_checksum in 7200s

Signed-off-by: Kent Overstreet <[email protected]>

show more ...

closures: closure_sync_timeout()

Add a new variant of closure_sync_timeout() that takes a timeout.

Note that when this returns -ETIME the closure will still be waiting on
something, i.e. it's not s

closures: closure_sync_timeout()

Add a new variant of closure_sync_timeout() that takes a timeout.

Note that when this returns -ETIME the closure will still be waiting on
something, i.e. it's not safe to return if you've got a stack allocated
closure.

Signed-off-by: Kent Overstreet <[email protected]>

show more ...

closures: CLOSURE_CALLBACK() to fix type punning

Control flow integrity is now checking that type signatures match on
indirect function calls. That breaks closures, which embed a work_struct
in a cl

closures: CLOSURE_CALLBACK() to fix type punning

Control flow integrity is now checking that type signatures match on
indirect function calls. That breaks closures, which embed a work_struct
in a closure in such a way that a closure_fn may also be used as a
workqueue fn by the underlying closure code.

So we have to change closure fns to take a work_struct as their
argument - but that results in a loss of clarity, as closure fns have
different semantics from normal workqueue functions (they run owning a
ref on the closure, which must be released with continue_at() or
closure_return()).

Thus, this patc introduces CLOSURE_CALLBACK() and closure_type() macros
as suggested by Kees, to smooth things over a bit.

Suggested-by: Kees Cook <[email protected]>
Cc: Coly Li <[email protected]>
Signed-off-by: Kent Overstreet <[email protected]>

show more ...

closures: Fix race in closure_sync()

As pointed out by Linus, closure_sync() was racy; we could skip blocking
immediately after a get() and a put(), but then that would skip any
barrier correspondin

closures: Fix race in closure_sync()

As pointed out by Linus, closure_sync() was racy; we could skip blocking
immediately after a get() and a put(), but then that would skip any
barrier corresponding to the other thread's put() barrier.

To fix this, always do the full __closure_sync() sequence whenever any
get() has happened and the closure might have been used by other
threads.

Signed-off-by: Kent Overstreet <[email protected]>

show more ...

closures: Better memory barriers

atomic_(dec|sub)_return_release() are a thing now - use them.

Also, delete the useless barrier in set_closure_fn(): it's redundant
with the memory barrier in closur

closures: Better memory barriers

atomic_(dec|sub)_return_release() are a thing now - use them.

Also, delete the useless barrier in set_closure_fn(): it's redundant
with the memory barrier in closure_put(0.

Since closure_put() would now otherwise just have a release barrier, we
also need a new barrier when the ref hits 0 -
smp_acquire__after_ctrl_dep().

Signed-off-by: Kent Overstreet <[email protected]>

show more ...

closures: Add a missing include

Fixes building in userspace.

Signed-off-by: Kent Overstreet <[email protected]>

bcache: move closures to lib/

Prep work for bcachefs - being a fork of bcache it also uses closures

Signed-off-by: Kent Overstreet <[email protected]>
Acked-by: Coly Li <[email protected]>
Rev

bcache: move closures to lib/

Prep work for bcachefs - being a fork of bcache it also uses closures

Signed-off-by: Kent Overstreet <[email protected]>
Acked-by: Coly Li <[email protected]>
Reviewed-by: Randy Dunlap <[email protected]>

show more ...