mirror of
https://github.com/AuxXxilium/linux_dsm_epyc7002.git
synced 2024-11-29 20:56:41 +07:00
0bce9c46bf
ARM recently moved to asm-generic/mutex-xchg.h for its mutex
implementation after the previous implementation was found to be missing
some crucial memory barriers. However, this has revealed some problems
running hackbench on SMP platforms due to the way in which the
MUTEX_SPIN_ON_OWNER code operates.
The symptoms are that a bunch of hackbench tasks are left waiting on an
unlocked mutex and therefore never get woken up to claim it. This boils
down to the following sequence of events:
Task A Task B Task C Lock value
0 1
1 lock() 0
2 lock() 0
3 spin(A) 0
4 unlock() 1
5 lock() 0
6 cmpxchg(1,0) 0
7 contended() -1
8 lock() 0
9 spin(C) 0
10 unlock() 1
11 cmpxchg(1,0) 0
12 unlock() 1
At this point, the lock is unlocked, but Task B is in an uninterruptible
sleep with nobody to wake it up.
This patch fixes the problem by ensuring we put the lock into the
contended state if we fail to acquire it on the fastpath, ensuring that
any blocked waiters are woken up when the mutex is released.
Signed-off-by: Will Deacon <will.deacon@arm.com>
Cc: Arnd Bergmann <arnd@arndb.de>
Cc: Chris Mason <chris.mason@fusionio.com>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: <stable@vger.kernel.org>
Reviewed-by: Nicolas Pitre <nico@linaro.org>
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Link: http://lkml.kernel.org/n/tip-6e9lrw2avczr0617fzl5vqb8@git.kernel.org
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
119 lines
3.9 KiB
C
119 lines
3.9 KiB
C
/*
|
|
* include/asm-generic/mutex-xchg.h
|
|
*
|
|
* Generic implementation of the mutex fastpath, based on xchg().
|
|
*
|
|
* NOTE: An xchg based implementation might be less optimal than an atomic
|
|
* decrement/increment based implementation. If your architecture
|
|
* has a reasonable atomic dec/inc then you should probably use
|
|
* asm-generic/mutex-dec.h instead, or you could open-code an
|
|
* optimized version in asm/mutex.h.
|
|
*/
|
|
#ifndef _ASM_GENERIC_MUTEX_XCHG_H
|
|
#define _ASM_GENERIC_MUTEX_XCHG_H
|
|
|
|
/**
|
|
* __mutex_fastpath_lock - try to take the lock by moving the count
|
|
* from 1 to a 0 value
|
|
* @count: pointer of type atomic_t
|
|
* @fail_fn: function to call if the original value was not 1
|
|
*
|
|
* Change the count from 1 to a value lower than 1, and call <fail_fn> if it
|
|
* wasn't 1 originally. This function MUST leave the value lower than 1
|
|
* even when the "1" assertion wasn't true.
|
|
*/
|
|
static inline void
|
|
__mutex_fastpath_lock(atomic_t *count, void (*fail_fn)(atomic_t *))
|
|
{
|
|
if (unlikely(atomic_xchg(count, 0) != 1))
|
|
/*
|
|
* We failed to acquire the lock, so mark it contended
|
|
* to ensure that any waiting tasks are woken up by the
|
|
* unlock slow path.
|
|
*/
|
|
if (likely(atomic_xchg(count, -1) != 1))
|
|
fail_fn(count);
|
|
}
|
|
|
|
/**
|
|
* __mutex_fastpath_lock_retval - try to take the lock by moving the count
|
|
* from 1 to a 0 value
|
|
* @count: pointer of type atomic_t
|
|
* @fail_fn: function to call if the original value was not 1
|
|
*
|
|
* Change the count from 1 to a value lower than 1, and call <fail_fn> if it
|
|
* wasn't 1 originally. This function returns 0 if the fastpath succeeds,
|
|
* or anything the slow path function returns
|
|
*/
|
|
static inline int
|
|
__mutex_fastpath_lock_retval(atomic_t *count, int (*fail_fn)(atomic_t *))
|
|
{
|
|
if (unlikely(atomic_xchg(count, 0) != 1))
|
|
if (likely(atomic_xchg(count, -1) != 1))
|
|
return fail_fn(count);
|
|
return 0;
|
|
}
|
|
|
|
/**
|
|
* __mutex_fastpath_unlock - try to promote the mutex from 0 to 1
|
|
* @count: pointer of type atomic_t
|
|
* @fail_fn: function to call if the original value was not 0
|
|
*
|
|
* try to promote the mutex from 0 to 1. if it wasn't 0, call <function>
|
|
* In the failure case, this function is allowed to either set the value to
|
|
* 1, or to set it to a value lower than one.
|
|
* If the implementation sets it to a value of lower than one, the
|
|
* __mutex_slowpath_needs_to_unlock() macro needs to return 1, it needs
|
|
* to return 0 otherwise.
|
|
*/
|
|
static inline void
|
|
__mutex_fastpath_unlock(atomic_t *count, void (*fail_fn)(atomic_t *))
|
|
{
|
|
if (unlikely(atomic_xchg(count, 1) != 0))
|
|
fail_fn(count);
|
|
}
|
|
|
|
#define __mutex_slowpath_needs_to_unlock() 0
|
|
|
|
/**
|
|
* __mutex_fastpath_trylock - try to acquire the mutex, without waiting
|
|
*
|
|
* @count: pointer of type atomic_t
|
|
* @fail_fn: spinlock based trylock implementation
|
|
*
|
|
* Change the count from 1 to a value lower than 1, and return 0 (failure)
|
|
* if it wasn't 1 originally, or return 1 (success) otherwise. This function
|
|
* MUST leave the value lower than 1 even when the "1" assertion wasn't true.
|
|
* Additionally, if the value was < 0 originally, this function must not leave
|
|
* it to 0 on failure.
|
|
*
|
|
* If the architecture has no effective trylock variant, it should call the
|
|
* <fail_fn> spinlock-based trylock variant unconditionally.
|
|
*/
|
|
static inline int
|
|
__mutex_fastpath_trylock(atomic_t *count, int (*fail_fn)(atomic_t *))
|
|
{
|
|
int prev = atomic_xchg(count, 0);
|
|
|
|
if (unlikely(prev < 0)) {
|
|
/*
|
|
* The lock was marked contended so we must restore that
|
|
* state. If while doing so we get back a prev value of 1
|
|
* then we just own it.
|
|
*
|
|
* [ In the rare case of the mutex going to 1, to 0, to -1
|
|
* and then back to 0 in this few-instructions window,
|
|
* this has the potential to trigger the slowpath for the
|
|
* owner's unlock path needlessly, but that's not a problem
|
|
* in practice. ]
|
|
*/
|
|
prev = atomic_xchg(count, prev);
|
|
if (prev < 0)
|
|
prev = 0;
|
|
}
|
|
|
|
return prev;
|
|
}
|
|
|
|
#endif
|