Brian Gerst wrote:
>
> resume_kernel uses CPU_IDX but never uses GET_CPU_IDX to get the index.
> This is an issue when smp and preemption are both enabled. I also
> removed the unused GET_CURRENT_CPU_IDX.
Gosh, your right. How did you find this?
I think the better solution, however, is to remove the whole
test. Since the preempt counter gets bumped on each
interrupt and every local_bh_disable, this code should NEVER
find either of the two counters other than zero (since we
already know preemt count is zero).
Good work.
-g
>
> --
> Brian Gerst
>
> ------------------------------------------------------------
> diff -urN linux-bk/arch/i386/kernel/entry.S linux/arch/i386/kernel/entry.S
> --- linux-bk/arch/i386/kernel/entry.S Wed May 29 15:06:00 2002
> +++ linux/arch/i386/kernel/entry.S Sat Jun 8 10:44:44 2002
> @@ -83,13 +83,9 @@
> #define GET_CPU_IDX \
> movl TI_CPU(%ebx), %eax; \
> shll $irq_array_shift, %eax
> -#define GET_CURRENT_CPU_IDX \
> - GET_THREAD_INFO(%ebx); \
> - GET_CPU_IDX
> #define CPU_IDX (,%eax)
> #else
> #define GET_CPU_IDX
> -#define GET_CURRENT_CPU_IDX GET_THREAD_INFO(%ebx)
> #define CPU_IDX
> #endif
>
> @@ -236,6 +232,7 @@
> movl TI_FLAGS(%ebx), %ecx
> testb $_TIF_NEED_RESCHED, %cl
> jz restore_all
> + GET_CPU_IDX
> movl irq_stat+local_bh_count CPU_IDX, %ecx
> addl irq_stat+local_irq_count CPU_IDX, %ecx
> jnz restore_all
-- George Anzinger george@mvista.com High-res-timers: http://sourceforge.net/projects/high-res-timers/ Real time sched: http://sourceforge.net/projects/rtsched/ Preemption patch: http://www.kernel.org/pub/linux/kernel/people/rml - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/
This archive was generated by hypermail 2b29 : Sat Jun 15 2002 - 22:00:13 EST