Re: [BUG] "sched: Remove rq->lock from the first half of ttwu()"locks up on ARM

[Yong Zhang]

On Fri, May 27, 2011 at 1:23 AM, Peter Zijlstra <peterz@xxxxxxxxxxxxx> wrote:
>
> We'd end up with something like the below, which isn't too different
> from what I've now got queued.
>
> It has the extra cpu == smp_processor_id() check, but I'm not sure this
> whole case is worth the trouble. I could go stick some counters in to
> verify how often all this happens I guess.
>
> ---
> Âarch/x86/include/asm/system.h | Â Â2 ++
> Âkernel/sched.c        Â|  14 +++++++++++---
> Âkernel/sched_debug.c     Â|  Â7 +++++++
> Â3 files changed, 20 insertions(+), 3 deletions(-)
>
> diff --git a/arch/x86/include/asm/system.h b/arch/x86/include/asm/system.h
> index c2ff2a1..2c597e8 100644
> --- a/arch/x86/include/asm/system.h
> +++ b/arch/x86/include/asm/system.h
> @@ -10,6 +10,8 @@
> Â#include <linux/kernel.h>
> Â#include <linux/irqflags.h>
>
> +#define __ARCH_WANT_INTERRUPTS_ON_CTXSW
> +
> Â/* entries in ARCH_DLINFO: */
> Â#if defined(CONFIG_IA32_EMULATION) || !defined(CONFIG_X86_64)
> Â# define AT_VECTOR_SIZE_ARCH 2
> diff --git a/kernel/sched.c b/kernel/sched.c
> index 2d12893..e4f7a9f 100644
> --- a/kernel/sched.c
> +++ b/kernel/sched.c
> @@ -2636,9 +2636,17 @@ try_to_wake_up(struct task_struct *p, unsigned int state, int wake_flags)
> Â Â Â Â Â Â Â Â * to spin on ->on_cpu if p is current, since that would
> Â Â Â Â Â Â Â Â * deadlock.
> Â Â Â Â Â Â Â Â */
> - Â Â Â Â Â Â Â if (p == current) {
> - Â Â Â Â Â Â Â Â Â Â Â ttwu_queue(p, cpu);
> - Â Â Â Â Â Â Â Â Â Â Â goto stat;
> + Â Â Â Â Â Â Â if (cpu == smp_processor_id()) {
> + Â Â Â Â Â Â Â Â Â Â Â struct rq *rq;
> +
> + Â Â Â Â Â Â Â Â Â Â Â rq = __task_rq_lock(p);
> + Â Â Â Â Â Â Â Â Â Â Â if (p->on_cpu) {

As Oleg has said, I also think we don't need this check.

> + Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â ttwu_activate(rq, p, ENQUEUE_WAKEUP);
> + Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â ttwu_do_wakeup(rq, p, wake_flags);

And the difference with ttwu_queue() is ttwu_queue() calls
ttwu_activate() with another flag ENQUEUE_WAKING, so if
we call ->task_waking() before  ttwu_queue(), I guess it will work
too.
But I like this version, because we call ->task_waking() and
ttwu_activate() on the local cpu, that means the calculations on
vruntime in that two functions are accumulated into noop.

Thanks,
Yong

> + Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â __task_rq_unlock(rq);
> + Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â goto stat;
> + Â Â Â Â Â Â Â Â Â Â Â }
> + Â Â Â Â Â Â Â Â Â Â Â __task_rq_unlock(rq);
> Â Â Â Â Â Â Â Â}
> Â#endif
> Â Â Â Â Â Â Â Âcpu_relax();
> diff --git a/kernel/sched_debug.c b/kernel/sched_debug.c
> index a6710a1..f0ff1de 100644
> --- a/kernel/sched_debug.c
> +++ b/kernel/sched_debug.c
> @@ -332,6 +332,13 @@ static int sched_debug_show(struct seq_file *m, void *v)
> Â Â Â Â Â Â Â Â(int)strcspn(init_utsname()->version, " "),
> Â Â Â Â Â Â Â Âinit_utsname()->version);
>
> +#ifdef __ARCH_WANT_INTERRUPTS_ON_CTXSW
> + Â Â Â SEQ_printf(m, "__ARCH_WANT_INTERRUPTS_ON_CTXSW\n");
> +#endif
> +#ifdef __ARCH_WANT_UNLOCKED_CTXSW
> + Â Â Â SEQ_printf(m, "__ARCH_WANT_UNLOCKED_CTXSW\n");
> +#endif
> +
> Â#define P(x) \
> Â Â Â ÂSEQ_printf(m, "%-40s: %Ld\n", #x, (long long)(x))
> Â#define PN(x) \
>
>
>



-- 
Only stand for myself
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/