Replace this pattern in osq_unlock():
atomic_cmpxchg(*ptr, old, new) == old
... with the simpler and faster:
atomic_try_cmpxchg(*ptr, &old, new)
The x86 CMPXCHG instruction returns success in the ZF flag,
so this change saves a compare after the CMPXCHG. The code
in the fast path of osq_unlock() improves from:
11b: 31 c9 xor %ecx,%ecx
11d: 8d 50 01 lea 0x1(%rax),%edx
120: 89 d0 mov %edx,%eax
122: f0 0f b1 0f lock cmpxchg %ecx,(%rdi)
126: 39 c2 cmp %eax,%edx
128: 75 05 jne 12f <...>
to:
12b: 31 d2 xor %edx,%edx
12d: 83 c0 01 add $0x1,%eax
130: f0 0f b1 17 lock cmpxchg %edx,(%rdi)
134: 75 05 jne 13b <...>
Signed-off-by: Uros Bizjak <ubizjak@gmail.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Acked-by: Waiman Long <longman@redhat.com>
Link: https://lore.kernel.org/r/20241001114606.820277-1-ubizjak@gmail.com
Every lock, that becomes a sleeping lock on PREEMPT_RT, starts a RCU read
side critical section. There is no sparse annotation for this and sparse
complains about unbalanced locking.
Add __acquires/ __releases for the RCU lock. This covers all but the
trylock functions. A __cond_acquires() annotation didn't work.
Signed-off-by: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Link: https://lore.kernel.org/all/20240812104200.2239232-4-bigeasy@linutronix.de
Add a function to check that an offline CPU has left the tracing
infrastructure in a sane state.
Commit 9bb69ba4c1 ("ACPI: processor_idle: use raw_safe_halt() in
acpi_idle_play_dead()") fixed an issue where the acpi_idle_play_dead()
function called safe_halt() instead of raw_safe_halt(), which had the
side-effect of setting the hardirqs_enabled flag for the offline CPU.
On x86 this triggered warnings from lockdep_assert_irqs_disabled() when
the CPU was brought back online again later. These warnings were too
early for the exception to be handled correctly, leading to a
triple-fault.
Add lockdep_cleanup_dead_cpu() to check for this kind of failure mode,
print the events leading up to it, and correct it so that the CPU can
come online again correctly. Re-introducing the original bug now merely
results in this warning instead:
[ 61.556652] smpboot: CPU 1 is now offline
[ 61.556769] CPU 1 left hardirqs enabled!
[ 61.556915] irq event stamp: 128149
[ 61.556965] hardirqs last enabled at (128149): [<ffffffff81720a36>] acpi_idle_play_dead+0x46/0x70
[ 61.557055] hardirqs last disabled at (128148): [<ffffffff81124d50>] do_idle+0x90/0xe0
[ 61.557117] softirqs last enabled at (128078): [<ffffffff81cec74c>] __do_softirq+0x31c/0x423
[ 61.557199] softirqs last disabled at (128065): [<ffffffff810baae1>] __irq_exit_rcu+0x91/0x100
[boqun: Capitalize the title and reword the message a bit]
Signed-off-by: David Woodhouse <dwmw@amazon.co.uk>
Reviewed-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Boqun Feng <boqun.feng@gmail.com>
Link: https://lore.kernel.org/r/f7bd2b3b999051bb3ef4be34526a9262008285f5.camel@infradead.org
With the proxy-execution series, we traverse the task->mutex->task
blocked_on/owner chain in the scheduler core. We do this while holding
the rq::lock to keep the structures in place while taking and
releasing the alternating lock types.
Since the mutex::wait_lock is one of the locks we will take in this
way under the rq::lock in the scheduler core, we need to make sure
that its usage elsewhere is irq safe.
[rebase & fix {un,}lock_wait_lock helpers in ww_mutex.h]
Signed-off-by: Juri Lelli <juri.lelli@redhat.com>
Signed-off-by: Connor O'Brien <connoro@google.com>
Signed-off-by: John Stultz <jstultz@google.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by: Metin Kaya <metin.kaya@arm.com>
Reviewed-by: Valentin Schneider <vschneid@redhat.com>
Tested-by: K Prateek Nayak <kprateek.nayak@amd.com>
Tested-by: Metin Kaya <metin.kaya@arm.com>
Link: https://lore.kernel.org/r/20241009235352.1614323-3-jstultz@google.com
When using mutex_acquire_nest() with a nest_lock, lockdep refcounts the
number of acquired lockdep_maps of mutexes of the same class, and also
keeps a pointer to the first acquired lockdep_map of a class. That pointer
is then used for various comparison-, printing- and checking purposes,
but there is no mechanism to actively ensure that lockdep_map stays in
memory. Instead, a warning is printed if the lockdep_map is freed and
there are still held locks of the same lock class, even if the lockdep_map
itself has been released.
In the context of WW/WD transactions that means that if a user unlocks
and frees a ww_mutex from within an ongoing ww transaction, and that
mutex happens to be the first ww_mutex grabbed in the transaction,
such a warning is printed and there might be a risk of a UAF.
Note that this is only problem when lockdep is enabled and affects only
dereferences of struct lockdep_map.
Adjust to this by adding a fake lockdep_map to the acquired context and
make sure it is the first acquired lockdep map of the associated
ww_mutex class. Then hold it for the duration of the WW/WD transaction.
This has the side effect that trying to lock a ww mutex *without* a
ww_acquire_context but where a such context has been acquire, we'd see
a lockdep splat. The test-ww_mutex.c selftest attempts to do that, so
modify that particular test to not acquire a ww_acquire_context if it
is not going to be used.
Signed-off-by: Thomas Hellström <thomas.hellstrom@linux.intel.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Link: https://lkml.kernel.org/r/20241009092031.6356-1-thomas.hellstrom@linux.intel.com
If CONFIG_GENERIC_LOCKBREAK=y and CONFIG_DEBUG_LOCK_ALLOC=n
(e.g. sh/sdk7786_defconfig):
kernel/locking/spinlock.c:68:17: warning: no previous prototype for '__raw_spin_lock' [-Wmissing-prototypes]
kernel/locking/spinlock.c:80:26: warning: no previous prototype for '__raw_spin_lock_irqsave' [-Wmissing-prototypes]
kernel/locking/spinlock.c:98:17: warning: no previous prototype for '__raw_spin_lock_irq' [-Wmissing-prototypes]
kernel/locking/spinlock.c:103:17: warning: no previous prototype for '__raw_spin_lock_bh' [-Wmissing-prototypes]
kernel/locking/spinlock.c:68:17: warning: no previous prototype for '__raw_read_lock' [-Wmissing-prototypes]
kernel/locking/spinlock.c:80:26: warning: no previous prototype for '__raw_read_lock_irqsave' [-Wmissing-prototypes]
kernel/locking/spinlock.c:98:17: warning: no previous prototype for '__raw_read_lock_irq' [-Wmissing-prototypes]
kernel/locking/spinlock.c:103:17: warning: no previous prototype for '__raw_read_lock_bh' [-Wmissing-prototypes]
kernel/locking/spinlock.c:68:17: warning: no previous prototype for '__raw_write_lock' [-Wmissing-prototypes]
kernel/locking/spinlock.c:80:26: warning: no previous prototype for '__raw_write_lock_irqsave' [-Wmissing-prototypes]
kernel/locking/spinlock.c:98:17: warning: no previous prototype for '__raw_write_lock_irq' [-Wmissing-prototypes]
kernel/locking/spinlock.c:103:17: warning: no previous prototype for '__raw_write_lock_bh' [-Wmissing-prototypes]
All __raw_* lock ops are internal functions without external callers.
Hence fix this by making them static.
Note that if CONFIG_GENERIC_LOCKBREAK=y, no lock ops are inlined, as all
of CONFIG_INLINE_*_LOCK* depend on !GENERIC_LOCKBREAK.
Signed-off-by: Geert Uytterhoeven <geert+renesas@glider.be>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Acked-by: Waiman Long <longman@redhat.com>
Link: https://lkml.kernel.org/r/7201d7fb408375c6c4df541270d787b1b4a32354.1727879348.git.geert+renesas@glider.be
Pull non-MM updates from Andrew Morton:
"Many singleton patches - please see the various changelogs for
details.
Quite a lot of nilfs2 work this time around.
Notable patch series in this pull request are:
- "mul_u64_u64_div_u64: new implementation" by Nicolas Pitre, with
assistance from Uwe Kleine-König. Reimplement mul_u64_u64_div_u64()
to provide (much) more accurate results. The current implementation
was causing Uwe some issues in the PWM drivers.
- "xz: Updates to license, filters, and compression options" from
Lasse Collin. Miscellaneous maintenance and kinor feature work to
the xz decompressor.
- "Fix some GDB command error and add some GDB commands" from
Kuan-Ying Lee. Fixes and enhancements to the gdb scripts.
- "treewide: add missing MODULE_DESCRIPTION() macros" from Jeff
Johnson. Adds lots of MODULE_DESCRIPTIONs, thus fixing lots of
warnings about this.
- "nilfs2: add support for some common ioctls" from Ryusuke Konishi.
Adds various commonly-available ioctls to nilfs2.
- "This series fixes a number of formatting issues in kernel doc
comments" from Ryusuke Konishi does that.
- "nilfs2: prevent unexpected ENOENT propagation" from Ryusuke
Konishi. Fix issues where -ENOENT was being unintentionally and
inappropriately returned to userspace.
- "nilfs2: assorted cleanups" from Huang Xiaojia.
- "nilfs2: fix potential issues with empty b-tree nodes" from Ryusuke
Konishi fixes some issues which can occur on corrupted nilfs2
filesystems.
- "scripts/decode_stacktrace.sh: improve error reporting and
usability" from Luca Ceresoli does those things"
* tag 'mm-nonmm-stable-2024-09-21-07-52' of git://git.kernel.org/pub/scm/linux/kernel/git/akpm/mm: (103 commits)
list: test: increase coverage of list_test_list_replace*()
list: test: fix tests for list_cut_position()
proc: use __auto_type more
treewide: correct the typo 'retun'
ocfs2: cleanup return value and mlog in ocfs2_global_read_info()
nilfs2: remove duplicate 'unlikely()' usage
nilfs2: fix potential oob read in nilfs_btree_check_delete()
nilfs2: determine empty node blocks as corrupted
nilfs2: fix potential null-ptr-deref in nilfs_btree_insert()
user_namespace: use kmemdup_array() instead of kmemdup() for multiple allocation
tools/mm: rm thp_swap_allocator_test when make clean
squashfs: fix percpu address space issues in decompressor_multi_percpu.c
lib: glob.c: added null check for character class
nilfs2: refactor nilfs_segctor_thread()
nilfs2: use kthread_create and kthread_stop for the log writer thread
nilfs2: remove sc_timer_task
nilfs2: do not repair reserved inode bitmap in nilfs_new_inode()
nilfs2: eliminate the shared counter and spinlock for i_generation
nilfs2: separate inode type information from i_state field
nilfs2: use the BITS_PER_LONG macro
...
Pull scheduler updates from Ingo Molnar:
- Implement the SCHED_DEADLINE server infrastructure - Daniel Bristot
de Oliveira's last major contribution to the kernel:
"SCHED_DEADLINE servers can help fixing starvation issues of low
priority tasks (e.g., SCHED_OTHER) when higher priority tasks
monopolize CPU cycles. Today we have RT Throttling; DEADLINE
servers should be able to replace and improve that."
(Daniel Bristot de Oliveira, Peter Zijlstra, Joel Fernandes, Youssef
Esmat, Huang Shijie)
- Preparatory changes for sched_ext integration:
- Use set_next_task(.first) where required
- Fix up set_next_task() implementations
- Clean up DL server vs. core sched
- Split up put_prev_task_balance()
- Rework pick_next_task()
- Combine the last put_prev_task() and the first set_next_task()
- Rework dl_server
- Add put_prev_task(.next)
(Peter Zijlstra, with a fix by Tejun Heo)
- Complete the EEVDF transition and refine EEVDF scheduling:
- Implement delayed dequeue
- Allow shorter slices to wakeup-preempt
- Use sched_attr::sched_runtime to set request/slice suggestion
- Document the new feature flags
- Remove unused and duplicate-functionality fields
- Simplify & unify pick_next_task_fair()
- Misc debuggability enhancements
(Peter Zijlstra, with fixes/cleanups by Dietmar Eggemann, Valentin
Schneider and Chuyi Zhou)
- Initialize the vruntime of a new task when it is first enqueued,
resulting in significant decrease in latency of newly woken tasks
(Zhang Qiao)
- Introduce SM_IDLE and an idle re-entry fast-path in __schedule()
(K Prateek Nayak, Peter Zijlstra)
- Clean up and clarify the usage of Clean up usage of rt_task()
(Qais Yousef)
- Preempt SCHED_IDLE entities in strict cgroup hierarchies
(Tianchen Ding)
- Clarify the documentation of time units for deadline scheduler
parameters (Christian Loehle)
- Remove the HZ_BW chicken-bit feature flag introduced a year ago,
the original change seems to be working fine (Phil Auld)
- Misc fixes and cleanups (Chen Yu, Dan Carpenter, Huang Shijie,
Peilin He, Qais Yousefm and Vincent Guittot)
* tag 'sched-core-2024-09-19' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip: (64 commits)
sched/cpufreq: Use NSEC_PER_MSEC for deadline task
cpufreq/cppc: Use NSEC_PER_MSEC for deadline task
sched/deadline: Clarify nanoseconds in uapi
sched/deadline: Convert schedtool example to chrt
sched/debug: Fix the runnable tasks output
sched: Fix sched_delayed vs sched_core
kernel/sched: Fix util_est accounting for DELAY_DEQUEUE
kthread: Fix task state in kthread worker if being frozen
sched/pelt: Use rq_clock_task() for hw_pressure
sched/fair: Move effective_cpu_util() and effective_cpu_util() in fair.c
sched/core: Introduce SM_IDLE and an idle re-entry fast-path in __schedule()
sched: Add put_prev_task(.next)
sched: Rework dl_server
sched: Combine the last put_prev_task() and the first set_next_task()
sched: Rework pick_next_task()
sched: Split up put_prev_task_balance()
sched: Clean up DL server vs core sched
sched: Fixup set_next_task() implementations
sched: Use set_next_task(.first) where required
sched/fair: Properly deactivate sched_delayed task upon class change
...
Pull printk updates from Petr Mladek:
"This is the "last" part of the support for the new nbcon consoles.
Where "nbcon" stays for "No Big console lock CONsoles" aka not under
the console_lock.
New callbacks are added to struct console:
- write_thread() for flushing nbcon consoles in task context.
- write_atomic() for flushing nbcon consoles in atomic context,
including NMI.
- con->device_lock() and device_unlock() for taking the driver
specific lock, for example, port->lock.
New printk-specific kthreads are created:
- per-console kthreads which get responsible for flushing normal
priority messages on nbcon consoles.
- thread which gets responsible for flushing normal priority messages
on all consoles when CONFIG_RT enabled.
The new callbacks are called under a special per-console lock which
has already been added back in v6.7. It allows to distinguish three
severities: normal, emergency, and panic. A context with a higher
priority could take over the ownership when it is safe even in the
middle of handling a record. The panic context could do it even when
it is not safe. But it is allowed only for the final desperate flush
before entering the infinite loop.
The new lock helps to flush the messages directly in emergency and
panic contexts. But it is not enough in all situations:
- console_lock() is still need for synchronization against boot
consoles.
- con->device_lock() is need for synchronization against other
operations on the same HW, e.g. serial port speed setting,
non-printk related read/write.
The dependency on con->device_lock() is mutual. Any code taking the
driver specific lock has to acquire the related nbcon console context
as well. For example, see the new uart_port_lock() API. It provides
the necessary synchronization against emergency and panic contexts
where the messages are flushed only under the new per-console lock.
Maybe surprisingly, a quite tricky part is the decision how to flush
the consoles in various situations. It has to take into account:
- message priority: normal, emergency, panic
- scheduling context: task, atomic, deferred_legacy
- registered consoles: boot, legacy, nbcon
- threads are running: early boot, suspend, shutdown, panic
- caller: printk(), pr_flush(), printk_flush_in_panic(),
console_unlock(), console_start(), ...
The primary decision is made in printk_get_console_flush_type(). It
creates a hint what the caller should do:
- flush nbcon consoles directly or via the kthread
- call the legacy loop (console_unlock()) directly or via irq_work
The existing behavior is preserved for the legacy consoles. The only
exception is that they are not longer flushed directly from printk()
in panic() before CPUs are stopped. But this blocking happens only
when at least one nbcon console is registered. The motivation is to
increase a chance to produce the crash dump. They legacy consoles
might create a deadlock in compare with nbcon consoles. The nbcon
console should allow to see the messages even when the crash dump
fails.
There are three possible ways how nbcon consoles are flushed:
- The per-nbcon-console kthread is responsible for flushing messages
added with the normal priority. This is the default mode.
- The legacy loop, aka console_unlock(), is used when there is still
a boot console registered. There is no easy way how to match an
early console driver with a nbcon console driver. And the
console_lock() provides the only reliable serialization at the
moment.
The legacy loop uses either con->write_atomic() or
con->write_thread() callbacks depending on whether it is allowed to
schedule. The atomic variant has to be used from printk().
- In other situations, the messages are flushed directly using
write_atomic() which can be called in any context, including NMI.
It is primary needed during early boot or shutdown, in emergency
situations, and panic.
The emergency priority is used by a code called within
nbcon_cpu_emergency_enter()/exit(). At the moment, it is used in four
situations: WARN(), Oops, lockdep, and RCU stall reports.
Finally, there is no nbcon console at the moment. It means that the
changes should _not_ modify the existing behavior. The only exception
is CONFIG_RT which would force offloading the legacy loop, for normal
priority context, into the dedicated kthread"
* tag 'printk-for-6.12' of git://git.kernel.org/pub/scm/linux/kernel/git/printk/linux: (54 commits)
printk: Avoid false positive lockdep report for legacy printing
printk: nbcon: Assign nice -20 for printing threads
printk: Implement legacy printer kthread for PREEMPT_RT
tty: sysfs: Add nbcon support for 'active'
proc: Add nbcon support for /proc/consoles
proc: consoles: Add notation to c_start/c_stop
printk: nbcon: Show replay message on takeover
printk: Provide helper for message prepending
printk: nbcon: Rely on kthreads for normal operation
printk: nbcon: Use thread callback if in task context for legacy
printk: nbcon: Relocate nbcon_atomic_emit_one()
printk: nbcon: Introduce printer kthreads
printk: nbcon: Init @nbcon_seq to highest possible
printk: nbcon: Add context to usable() and emit()
printk: Flush console on unregister_console()
printk: Fail pr_flush() if before SYSTEM_SCHEDULING
printk: nbcon: Add function for printers to reacquire ownership
printk: nbcon: Use raw_cpu_ptr() instead of open coding
printk: Use the BITS_PER_LONG macro
lockdep: Mark emergency sections in lockdep splats
...
Both is_rwsem_reader_owned() and rwsem_owner() are currently only used when
CONFIG_DEBUG_RWSEMS is defined. This causes a compilation error with clang
when `make W=1` and CONFIG_WERROR=y:
kernel/locking/rwsem.c:187:20: error: unused function 'is_rwsem_reader_owned' [-Werror,-Wunused-function]
187 | static inline bool is_rwsem_reader_owned(struct rw_semaphore *sem)
| ^~~~~~~~~~~~~~~~~~~~~
kernel/locking/rwsem.c:271:35: error: unused function 'rwsem_owner' [-Werror,-Wunused-function]
271 | static inline struct task_struct *rwsem_owner(struct rw_semaphore *sem)
| ^~~~~~~~~~~
Fix this by moving these two functions under the CONFIG_DEBUG_RWSEMS define.
Reported-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Signed-off-by: Waiman Long <longman@redhat.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Tested-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Link: https://lore.kernel.org/r/20240909182905.161156-1-longman@redhat.com
Mark emergency sections wherever multiple lines of
lock debugging output are generated. In an emergency
section, every printk() call will attempt to directly
flush to the consoles using the EMERGENCY priority.
Note that debug_show_all_locks() and
lockdep_print_held_locks() rely on their callers to
enter the emergency section. This is because these
functions can also be called in non-emergency
situations (such as sysrq).
Signed-off-by: John Ogness <john.ogness@linutronix.de>
Reviewed-by: Petr Mladek <pmladek@suse.com>
Link: https://lore.kernel.org/r/20240820063001.36405-36-john.ogness@linutronix.de
Signed-off-by: Petr Mladek <pmladek@suse.com>
rt_mutex_handle_deadlock() is called with rt_mutex::wait_lock held. In the
good case it returns with the lock held and in the deadlock case it emits a
warning and goes into an endless scheduling loop with the lock held, which
triggers the 'scheduling in atomic' warning.
Unlock rt_mutex::wait_lock in the dead lock case before issuing the warning
and dropping into the schedule for ever loop.
[ tglx: Moved unlock before the WARN(), removed the pointless comment,
massaged changelog, added Fixes tag ]
Fixes: 3d5c9340d1 ("rtmutex: Handle deadlock detection smarter")
Signed-off-by: Roland Xu <mu001999@outlook.com>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Cc: stable@vger.kernel.org
Link: https://lore.kernel.org/all/ME0P300MB063599BEF0743B8FA339C2CECC802@ME0P300MB0635.AUSP300.PROD.OUTLOOK.COM
Pull bcachefs fixes from Kent Overstreet:
"Assorted little stuff:
- lockdep fixup for lockdep_set_notrack_class()
- we can now remove a device when using erasure coding without
deadlocking, though we still hit other issues
- the 'allocator stuck' timeout is now configurable, and messages are
ratelimited. The default timeout has been increased from 10 seconds
to 30"
* tag 'bcachefs-2024-08-08' of git://evilpiepirate.org/bcachefs:
bcachefs: Use bch2_wait_on_allocator() in btree node alloc path
bcachefs: Make allocator stuck timeout configurable, ratelimit messages
bcachefs: Add missing path_traverse() to btree_iter_next_node()
bcachefs: ec should not allocate from ro devs
bcachefs: Improved allocator debugging for ec
bcachefs: Add missing bch2_trans_begin() call
bcachefs: Add a comment for bucket helper types
bcachefs: Don't rely on implicit unsigned -> signed integer conversion
lockdep: Fix lockdep_set_notrack_class() for CONFIG_LOCK_STAT
bcachefs: Fix double free of ca->buckets_nouse
rt_task() checks if a task has RT priority. But depends on your
dictionary, this could mean it belongs to RT class, or is a 'realtime'
task, which includes RT and DL classes.
Since this has caused some confusion already on discussion [1], it
seemed a clean up is due.
I define the usage of rt_task() to be tasks that belong to RT class.
Make sure that it returns true only for RT class and audit the users and
replace the ones required the old behavior with the new realtime_task()
which returns true for RT and DL classes. Introduce similar
realtime_prio() to create similar distinction to rt_prio() and update
the users that required the old behavior to use the new function.
Move MAX_DL_PRIO to prio.h so it can be used in the new definitions.
Document the functions to make it more obvious what is the difference
between them. PI-boosted tasks is a factor that must be taken into
account when choosing which function to use.
Rename task_is_realtime() to realtime_task_policy() as the old name is
confusing against the new realtime_task().
No functional changes were intended.
[1] https://lore.kernel.org/lkml/20240506100509.GL40213@noisy.programming.kicks-ass.net/
Signed-off-by: Qais Yousef <qyousef@layalina.io>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by: Phil Auld <pauld@redhat.com>
Reviewed-by: "Steven Rostedt (Google)" <rostedt@goodmis.org>
Reviewed-by: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Link: https://lore.kernel.org/r/20240610192018.1567075-2-qyousef@layalina.io
