mirror of
https://github.com/AuxXxilium/linux_dsm_epyc7002.git
synced 2024-11-30 18:36:39 +07:00
ef6edc9746
On systems running with virtual cpus there is optimization potential in regard to spinlocks and rw-locks. If the virtual cpu that has taken a lock is known to a cpu that wants to acquire the same lock it is beneficial to yield the timeslice of the virtual cpu in favour of the cpu that has the lock (directed yield). With CONFIG_PREEMPT="n" this can be implemented by the architecture without common code changes. Powerpc already does this. With CONFIG_PREEMPT="y" the lock loops are coded with _raw_spin_trylock, _raw_read_trylock and _raw_write_trylock in kernel/spinlock.c. If the lock could not be taken cpu_relax is called. A directed yield is not possible because cpu_relax doesn't know anything about the lock. To be able to yield the lock in favour of the current lock holder variants of cpu_relax for spinlocks and rw-locks are needed. The new _raw_spin_relax, _raw_read_relax and _raw_write_relax primitives differ from cpu_relax insofar that they have an argument: a pointer to the lock structure. Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com> Cc: Ingo Molnar <mingo@elte.hu> Cc: Paul Mackerras <paulus@samba.org> Cc: Haavard Skinnemoen <hskinnemoen@atmel.com> Signed-off-by: Andrew Morton <akpm@osdl.org> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
324 lines
7.6 KiB
C
324 lines
7.6 KiB
C
#ifndef _ASM_M32R_SPINLOCK_H
|
|
#define _ASM_M32R_SPINLOCK_H
|
|
|
|
/*
|
|
* linux/include/asm-m32r/spinlock.h
|
|
*
|
|
* M32R version:
|
|
* Copyright (C) 2001, 2002 Hitoshi Yamamoto
|
|
* Copyright (C) 2004 Hirokazu Takata <takata at linux-m32r.org>
|
|
*/
|
|
|
|
#include <linux/compiler.h>
|
|
#include <asm/atomic.h>
|
|
#include <asm/page.h>
|
|
|
|
/*
|
|
* Your basic SMP spinlocks, allowing only a single CPU anywhere
|
|
*
|
|
* (the type definitions are in asm/spinlock_types.h)
|
|
*
|
|
* Simple spin lock operations. There are two variants, one clears IRQ's
|
|
* on the local processor, one does not.
|
|
*
|
|
* We make no fairness assumptions. They have a cost.
|
|
*/
|
|
|
|
#define __raw_spin_is_locked(x) (*(volatile int *)(&(x)->slock) <= 0)
|
|
#define __raw_spin_lock_flags(lock, flags) __raw_spin_lock(lock)
|
|
#define __raw_spin_unlock_wait(x) \
|
|
do { cpu_relax(); } while (__raw_spin_is_locked(x))
|
|
|
|
/**
|
|
* __raw_spin_trylock - Try spin lock and return a result
|
|
* @lock: Pointer to the lock variable
|
|
*
|
|
* __raw_spin_trylock() tries to get the lock and returns a result.
|
|
* On the m32r, the result value is 1 (= Success) or 0 (= Failure).
|
|
*/
|
|
static inline int __raw_spin_trylock(raw_spinlock_t *lock)
|
|
{
|
|
int oldval;
|
|
unsigned long tmp1, tmp2;
|
|
|
|
/*
|
|
* lock->slock : =1 : unlock
|
|
* : <=0 : lock
|
|
* {
|
|
* oldval = lock->slock; <--+ need atomic operation
|
|
* lock->slock = 0; <--+
|
|
* }
|
|
*/
|
|
__asm__ __volatile__ (
|
|
"# __raw_spin_trylock \n\t"
|
|
"ldi %1, #0; \n\t"
|
|
"mvfc %2, psw; \n\t"
|
|
"clrpsw #0x40 -> nop; \n\t"
|
|
DCACHE_CLEAR("%0", "r6", "%3")
|
|
"lock %0, @%3; \n\t"
|
|
"unlock %1, @%3; \n\t"
|
|
"mvtc %2, psw; \n\t"
|
|
: "=&r" (oldval), "=&r" (tmp1), "=&r" (tmp2)
|
|
: "r" (&lock->slock)
|
|
: "memory"
|
|
#ifdef CONFIG_CHIP_M32700_TS1
|
|
, "r6"
|
|
#endif /* CONFIG_CHIP_M32700_TS1 */
|
|
);
|
|
|
|
return (oldval > 0);
|
|
}
|
|
|
|
static inline void __raw_spin_lock(raw_spinlock_t *lock)
|
|
{
|
|
unsigned long tmp0, tmp1;
|
|
|
|
/*
|
|
* lock->slock : =1 : unlock
|
|
* : <=0 : lock
|
|
*
|
|
* for ( ; ; ) {
|
|
* lock->slock -= 1; <-- need atomic operation
|
|
* if (lock->slock == 0) break;
|
|
* for ( ; lock->slock <= 0 ; );
|
|
* }
|
|
*/
|
|
__asm__ __volatile__ (
|
|
"# __raw_spin_lock \n\t"
|
|
".fillinsn \n"
|
|
"1: \n\t"
|
|
"mvfc %1, psw; \n\t"
|
|
"clrpsw #0x40 -> nop; \n\t"
|
|
DCACHE_CLEAR("%0", "r6", "%2")
|
|
"lock %0, @%2; \n\t"
|
|
"addi %0, #-1; \n\t"
|
|
"unlock %0, @%2; \n\t"
|
|
"mvtc %1, psw; \n\t"
|
|
"bltz %0, 2f; \n\t"
|
|
LOCK_SECTION_START(".balign 4 \n\t")
|
|
".fillinsn \n"
|
|
"2: \n\t"
|
|
"ld %0, @%2; \n\t"
|
|
"bgtz %0, 1b; \n\t"
|
|
"bra 2b; \n\t"
|
|
LOCK_SECTION_END
|
|
: "=&r" (tmp0), "=&r" (tmp1)
|
|
: "r" (&lock->slock)
|
|
: "memory"
|
|
#ifdef CONFIG_CHIP_M32700_TS1
|
|
, "r6"
|
|
#endif /* CONFIG_CHIP_M32700_TS1 */
|
|
);
|
|
}
|
|
|
|
static inline void __raw_spin_unlock(raw_spinlock_t *lock)
|
|
{
|
|
mb();
|
|
lock->slock = 1;
|
|
}
|
|
|
|
/*
|
|
* Read-write spinlocks, allowing multiple readers
|
|
* but only one writer.
|
|
*
|
|
* NOTE! it is quite common to have readers in interrupts
|
|
* but no interrupt writers. For those circumstances we
|
|
* can "mix" irq-safe locks - any writer needs to get a
|
|
* irq-safe write-lock, but readers can get non-irqsafe
|
|
* read-locks.
|
|
*
|
|
* On x86, we implement read-write locks as a 32-bit counter
|
|
* with the high bit (sign) being the "contended" bit.
|
|
*
|
|
* The inline assembly is non-obvious. Think about it.
|
|
*
|
|
* Changed to use the same technique as rw semaphores. See
|
|
* semaphore.h for details. -ben
|
|
*/
|
|
|
|
/**
|
|
* read_can_lock - would read_trylock() succeed?
|
|
* @lock: the rwlock in question.
|
|
*/
|
|
#define __raw_read_can_lock(x) ((int)(x)->lock > 0)
|
|
|
|
/**
|
|
* write_can_lock - would write_trylock() succeed?
|
|
* @lock: the rwlock in question.
|
|
*/
|
|
#define __raw_write_can_lock(x) ((x)->lock == RW_LOCK_BIAS)
|
|
|
|
static inline void __raw_read_lock(raw_rwlock_t *rw)
|
|
{
|
|
unsigned long tmp0, tmp1;
|
|
|
|
/*
|
|
* rw->lock : >0 : unlock
|
|
* : <=0 : lock
|
|
*
|
|
* for ( ; ; ) {
|
|
* rw->lock -= 1; <-- need atomic operation
|
|
* if (rw->lock >= 0) break;
|
|
* rw->lock += 1; <-- need atomic operation
|
|
* for ( ; rw->lock <= 0 ; );
|
|
* }
|
|
*/
|
|
__asm__ __volatile__ (
|
|
"# read_lock \n\t"
|
|
".fillinsn \n"
|
|
"1: \n\t"
|
|
"mvfc %1, psw; \n\t"
|
|
"clrpsw #0x40 -> nop; \n\t"
|
|
DCACHE_CLEAR("%0", "r6", "%2")
|
|
"lock %0, @%2; \n\t"
|
|
"addi %0, #-1; \n\t"
|
|
"unlock %0, @%2; \n\t"
|
|
"mvtc %1, psw; \n\t"
|
|
"bltz %0, 2f; \n\t"
|
|
LOCK_SECTION_START(".balign 4 \n\t")
|
|
".fillinsn \n"
|
|
"2: \n\t"
|
|
"clrpsw #0x40 -> nop; \n\t"
|
|
DCACHE_CLEAR("%0", "r6", "%2")
|
|
"lock %0, @%2; \n\t"
|
|
"addi %0, #1; \n\t"
|
|
"unlock %0, @%2; \n\t"
|
|
"mvtc %1, psw; \n\t"
|
|
".fillinsn \n"
|
|
"3: \n\t"
|
|
"ld %0, @%2; \n\t"
|
|
"bgtz %0, 1b; \n\t"
|
|
"bra 3b; \n\t"
|
|
LOCK_SECTION_END
|
|
: "=&r" (tmp0), "=&r" (tmp1)
|
|
: "r" (&rw->lock)
|
|
: "memory"
|
|
#ifdef CONFIG_CHIP_M32700_TS1
|
|
, "r6"
|
|
#endif /* CONFIG_CHIP_M32700_TS1 */
|
|
);
|
|
}
|
|
|
|
static inline void __raw_write_lock(raw_rwlock_t *rw)
|
|
{
|
|
unsigned long tmp0, tmp1, tmp2;
|
|
|
|
/*
|
|
* rw->lock : =RW_LOCK_BIAS_STR : unlock
|
|
* : !=RW_LOCK_BIAS_STR : lock
|
|
*
|
|
* for ( ; ; ) {
|
|
* rw->lock -= RW_LOCK_BIAS_STR; <-- need atomic operation
|
|
* if (rw->lock == 0) break;
|
|
* rw->lock += RW_LOCK_BIAS_STR; <-- need atomic operation
|
|
* for ( ; rw->lock != RW_LOCK_BIAS_STR ; ) ;
|
|
* }
|
|
*/
|
|
__asm__ __volatile__ (
|
|
"# write_lock \n\t"
|
|
"seth %1, #high(" RW_LOCK_BIAS_STR "); \n\t"
|
|
"or3 %1, %1, #low(" RW_LOCK_BIAS_STR "); \n\t"
|
|
".fillinsn \n"
|
|
"1: \n\t"
|
|
"mvfc %2, psw; \n\t"
|
|
"clrpsw #0x40 -> nop; \n\t"
|
|
DCACHE_CLEAR("%0", "r7", "%3")
|
|
"lock %0, @%3; \n\t"
|
|
"sub %0, %1; \n\t"
|
|
"unlock %0, @%3; \n\t"
|
|
"mvtc %2, psw; \n\t"
|
|
"bnez %0, 2f; \n\t"
|
|
LOCK_SECTION_START(".balign 4 \n\t")
|
|
".fillinsn \n"
|
|
"2: \n\t"
|
|
"clrpsw #0x40 -> nop; \n\t"
|
|
DCACHE_CLEAR("%0", "r7", "%3")
|
|
"lock %0, @%3; \n\t"
|
|
"add %0, %1; \n\t"
|
|
"unlock %0, @%3; \n\t"
|
|
"mvtc %2, psw; \n\t"
|
|
".fillinsn \n"
|
|
"3: \n\t"
|
|
"ld %0, @%3; \n\t"
|
|
"beq %0, %1, 1b; \n\t"
|
|
"bra 3b; \n\t"
|
|
LOCK_SECTION_END
|
|
: "=&r" (tmp0), "=&r" (tmp1), "=&r" (tmp2)
|
|
: "r" (&rw->lock)
|
|
: "memory"
|
|
#ifdef CONFIG_CHIP_M32700_TS1
|
|
, "r7"
|
|
#endif /* CONFIG_CHIP_M32700_TS1 */
|
|
);
|
|
}
|
|
|
|
static inline void __raw_read_unlock(raw_rwlock_t *rw)
|
|
{
|
|
unsigned long tmp0, tmp1;
|
|
|
|
__asm__ __volatile__ (
|
|
"# read_unlock \n\t"
|
|
"mvfc %1, psw; \n\t"
|
|
"clrpsw #0x40 -> nop; \n\t"
|
|
DCACHE_CLEAR("%0", "r6", "%2")
|
|
"lock %0, @%2; \n\t"
|
|
"addi %0, #1; \n\t"
|
|
"unlock %0, @%2; \n\t"
|
|
"mvtc %1, psw; \n\t"
|
|
: "=&r" (tmp0), "=&r" (tmp1)
|
|
: "r" (&rw->lock)
|
|
: "memory"
|
|
#ifdef CONFIG_CHIP_M32700_TS1
|
|
, "r6"
|
|
#endif /* CONFIG_CHIP_M32700_TS1 */
|
|
);
|
|
}
|
|
|
|
static inline void __raw_write_unlock(raw_rwlock_t *rw)
|
|
{
|
|
unsigned long tmp0, tmp1, tmp2;
|
|
|
|
__asm__ __volatile__ (
|
|
"# write_unlock \n\t"
|
|
"seth %1, #high(" RW_LOCK_BIAS_STR "); \n\t"
|
|
"or3 %1, %1, #low(" RW_LOCK_BIAS_STR "); \n\t"
|
|
"mvfc %2, psw; \n\t"
|
|
"clrpsw #0x40 -> nop; \n\t"
|
|
DCACHE_CLEAR("%0", "r7", "%3")
|
|
"lock %0, @%3; \n\t"
|
|
"add %0, %1; \n\t"
|
|
"unlock %0, @%3; \n\t"
|
|
"mvtc %2, psw; \n\t"
|
|
: "=&r" (tmp0), "=&r" (tmp1), "=&r" (tmp2)
|
|
: "r" (&rw->lock)
|
|
: "memory"
|
|
#ifdef CONFIG_CHIP_M32700_TS1
|
|
, "r7"
|
|
#endif /* CONFIG_CHIP_M32700_TS1 */
|
|
);
|
|
}
|
|
|
|
static inline int __raw_read_trylock(raw_rwlock_t *lock)
|
|
{
|
|
atomic_t *count = (atomic_t*)lock;
|
|
if (atomic_dec_return(count) >= 0)
|
|
return 1;
|
|
atomic_inc(count);
|
|
return 0;
|
|
}
|
|
|
|
static inline int __raw_write_trylock(raw_rwlock_t *lock)
|
|
{
|
|
atomic_t *count = (atomic_t *)lock;
|
|
if (atomic_sub_and_test(RW_LOCK_BIAS, count))
|
|
return 1;
|
|
atomic_add(RW_LOCK_BIAS, count);
|
|
return 0;
|
|
}
|
|
|
|
#define _raw_spin_relax(lock) cpu_relax()
|
|
#define _raw_read_relax(lock) cpu_relax()
|
|
#define _raw_write_relax(lock) cpu_relax()
|
|
|
|
#endif /* _ASM_M32R_SPINLOCK_H */
|