Re: [PATCH] MIPS: asm: spinlock: Adjust arch_spin_lock back-off time

[David Daney <ddaney.cavm@gmail.com>]

On 04/20/2015 02:54 AM, Markos Chandras wrote:
> Make it similar to the trylock and R10000_LLSC_WAR cases.
>
> Signed-off-by: Markos Chandras <markos.chandras@imgtec.com>
> ---
> I can't find a reference on why this was different compared to the other
> cases so I presume it was just an overlook in commit
> 2a31b03335e570dce5fdd082e0d71d48b2cb4290

I suspect you are correct.  Although in practice, it doesn't matter.

We are doing a mask of the difference between the ticket and now serving 
values.  You would have to have more than 0x1fff (8191) CPUs in a system 
to hit this limit.

That said, I think the patch is OK.

David Daney


> ---
>   arch/mips/include/asm/spinlock.h | 2 +-
>   1 file changed, 1 insertion(+), 1 deletion(-)
>
> diff --git a/arch/mips/include/asm/spinlock.h b/arch/mips/include/asm/spinlock.h
> index 1fca2e0793dc..cc771a21899d 100644
> --- a/arch/mips/include/asm/spinlock.h
> +++ b/arch/mips/include/asm/spinlock.h
> @@ -109,7 +109,7 @@ static inline void arch_spin_lock(arch_spinlock_t *lock)
>   		"	 subu	%[ticket], %[my_ticket], %[ticket]	\n"
>   		"2:							\n"
>   		"	.subsection 2					\n"
> -		"4:	andi	%[ticket], %[ticket], 0x1fff		\n"
> +		"4:	andi	%[ticket], %[ticket], 0xffff		\n"
>   		"	sll	%[ticket], 5				\n"
>   		"							\n"
>   		"6:	bnez	%[ticket], 6b				\n"
>