Re: [PATCH v3 10/25] openrisc: add spinlock implementation
From: Peter Zijlstra
Date: Wed Feb 22 2017 - 06:37:32 EST
On Wed, Feb 22, 2017 at 04:11:39AM +0900, Stafford Horne wrote:
> +static inline int arch_spin_trylock(arch_spinlock_t *lock)
> +{
> + unsigned long contended, tmp;
> + u32 slock;
> +
> + /* contended = (lock->tickets.owner != lock->tickets.next) */
> + __asm__ __volatile__(
> + "1: l.lwa %0, 0(%3) \n"
> + " l.srli %1, %0, 16 \n"
> + " l.andi %2, %0, 0xffff \n"
> + " l.sfeq %1, %2 \n"
> + " l.bnf 1f \n"
> + " l.ori %1, r0, 1 \n"
> + " l.add %0, %0, %4 \n"
> + " l.swa 0(%3), %0 \n"
> + " l.bnf 1b \n"
> + " l.ori %1, r0, 0 \n"
#ifdef CONFIG_SMP
" l.sync \n"
#endif
> + "1: \n"
> + : "=&r" (slock), "=&r" (contended), "=&r" (tmp)
> + : "r" (&lock->slock), "r" (1 << TICKET_SHIFT)
> + : "cc", "memory");
> +
Then s/contended/acquired/, flip the bitset in the asm, and replace the
entire thing below with:
return acquired;
> + if (!contended) {
> + smp_mb();
> + return 1;
> + } else {
> + return 0;
> + }
> +}