rust: helpers: Avoid raw_spin_lock initialization for PREEMPT_RT

When PREEMPT_RT=y, spin locks are mapped to rt_mutex types, so using
spinlock_check() + __raw_spin_lock_init() to initialize spin loc

rust: helpers: Avoid raw_spin_lock initialization for PREEMPT_RT

When PREEMPT_RT=y, spin locks are mapped to rt_mutex types, so using
spinlock_check() + __raw_spin_lock_init() to initialize spin locks is
incorrect, and would cause build errors.

Introduce __spin_lock_init() to initialize a spin lock with lockdep
rquired information for PREEMPT_RT builds, and use it in the Rust
helper.

Fixes: d2d6422f8bd1 ("x86: Allow to enable PREEMPT_RT.")
Closes: https://lore.kernel.org/oe-kbuild-all/[email protected]/
Reported-by: kernel test robot <[email protected]>
Signed-off-by: Eder Zulian <[email protected]>
Signed-off-by: Peter Zijlstra (Intel) <[email protected]>
Reviewed-by: Boqun Feng <[email protected]>
Tested-by: Boqun Feng <[email protected]>
Link: https://lore.kernel.org/r/[email protected]

show more ...

locking/rt: Remove one __cond_lock() in RT's spin_trylock_irqsave()

spin_trylock_irqsave() has a __cond_lock() wrapper which points to
__spin_trylock_irqsave(). The function then invokes spin_tryloc

locking/rt: Remove one __cond_lock() in RT's spin_trylock_irqsave()

spin_trylock_irqsave() has a __cond_lock() wrapper which points to
__spin_trylock_irqsave(). The function then invokes spin_trylock() which
has another __cond_lock() finally pointing to rt_spin_trylock().

The compiler has no problem to parse this but sparse does not recognise
that users of spin_trylock_irqsave() acquire a conditional lock and
complains.

Remove one layer of __cond_lock() so that sparse recognises conditional
locking.

Signed-off-by: Sebastian Andrzej Siewior <[email protected]>
Signed-off-by: Thomas Gleixner <[email protected]>
Link: https://lore.kernel.org/all/[email protected]

show more ...

locking/rt: Add sparse annotation PREEMPT_RT's sleeping locks.

The sleeping locks on PREEMPT_RT (rt_spin_lock() and friends) lack
sparse annotation. Therefore a missing spin_unlock() won't be spotte

locking/rt: Add sparse annotation PREEMPT_RT's sleeping locks.

The sleeping locks on PREEMPT_RT (rt_spin_lock() and friends) lack
sparse annotation. Therefore a missing spin_unlock() won't be spotted by
sparse in a PREEMPT_RT build while it is noticed on a !PREEMPT_RT build.

Add the __acquires/__releases macros to the lock/ unlock functions. The
trylock functions already use the __cond_lock() wrapper.

Signed-off-by: Sebastian Andrzej Siewior <[email protected]>
Signed-off-by: Thomas Gleixner <[email protected]>
Link: https://lore.kernel.org/all/[email protected]

show more ...

locking: Detect includes rwlock.h outside of spinlock.h

From: Michael S. Tsirkin <[email protected]>

The check for __LINUX_SPINLOCK_H within rwlock.h (and other files)
detects the direct include of th

locking: Detect includes rwlock.h outside of spinlock.h

From: Michael S. Tsirkin <[email protected]>

The check for __LINUX_SPINLOCK_H within rwlock.h (and other files)
detects the direct include of the header file if it is at the very
beginning of the include section.
If it is listed later then chances are high that spinlock.h was already
included (including rwlock.h) and the additional listing of rwlock.h
will not cause any failure.

On PREEMPT_RT this additional rwlock.h will lead to compile failures
since it uses a different rwlock implementation.

Add __LINUX_INSIDE_SPINLOCK_H to spinlock.h and check for this instead
of __LINUX_SPINLOCK_H to detect wrong includes. This will help detect
direct includes of rwlock.h with without PREEMPT_RT enabled.

[ bigeasy: add remaining __LINUX_SPINLOCK_H user and rewrite
commit description. ]

Signed-off-by: Michael S. Tsirkin <[email protected]>
Signed-off-by: Sebastian Andrzej Siewior <[email protected]>
Signed-off-by: Peter Zijlstra (Intel) <[email protected]>
Link: https://lkml.kernel.org/r/[email protected]

show more ...

locking/spinlock/rt: Prepare for RT local_lock

Add the static and runtime initializer mechanics to support the RT variant
of local_lock, which requires the lock type in the lockdep map to be set
to

locking/spinlock/rt: Prepare for RT local_lock

Add the static and runtime initializer mechanics to support the RT variant
of local_lock, which requires the lock type in the lockdep map to be set
to LD_LOCK_PERCPU.

Signed-off-by: Thomas Gleixner <[email protected]>
Signed-off-by: Peter Zijlstra (Intel) <[email protected]>
Signed-off-by: Ingo Molnar <[email protected]>
Link: https://lore.kernel.org/r/[email protected]

show more ...

locking/rwlock: Provide RT variant

Similar to rw_semaphores, on RT the rwlock substitution is not writer fair,
because it's not feasible to have a writer inherit its priority to
multiple readers. Re

locking/rwlock: Provide RT variant

Similar to rw_semaphores, on RT the rwlock substitution is not writer fair,
because it's not feasible to have a writer inherit its priority to
multiple readers. Readers blocked on a writer follow the normal rules of
priority inheritance. Like RT spinlocks, RT rwlocks are state preserving
across the slow lock operations (contended case).

Signed-off-by: Thomas Gleixner <[email protected]>
Signed-off-by: Peter Zijlstra (Intel) <[email protected]>
Signed-off-by: Ingo Molnar <[email protected]>
Link: https://lore.kernel.org/r/[email protected]

show more ...

locking/spinlock: Provide RT variant header: <linux/spinlock_rt.h>

Provide the necessary wrappers around the actual rtmutex based spinlock
implementation.

Signed-off-by: Thomas Gleixner <tglx@linut

locking/spinlock: Provide RT variant header: <linux/spinlock_rt.h>

Provide the necessary wrappers around the actual rtmutex based spinlock
implementation.

Signed-off-by: Thomas Gleixner <[email protected]>
Signed-off-by: Peter Zijlstra (Intel) <[email protected]>
Signed-off-by: Ingo Molnar <[email protected]>
Link: https://lore.kernel.org/r/[email protected]

show more ...