[PATCH tip/core/rcu 1/3] membarrier: Provide register expedited private command

[Nicholas Piggin]

On Wed,  4 Oct 2017 14:37:53 -0700
"Paul E. McKenney" <paulmck at linux.vnet.ibm.com> wrote:

> From: Mathieu Desnoyers <mathieu.desnoyers at efficios.com>
> 
> Provide a new command allowing processes to register their intent to use
> the private expedited command.
> 
> This allows PowerPC to skip the full memory barrier in switch_mm(), and
> only issue the barrier when scheduling into a task belonging to a
> process that has registered to use expedited private.
> 
> Processes are now required to register before using
> MEMBARRIER_CMD_PRIVATE_EXPEDITED, otherwise that command returns EPERM.
> 
> Changes since v1:
> - Use test_ti_thread_flag(next, ...) instead of test_thread_flag() in
>   powerpc membarrier_arch_sched_in(), given that we want to specifically
>   check the next thread state.
> - Add missing ARCH_HAS_MEMBARRIER_HOOKS in Kconfig.
> - Use task_thread_info() to pass thread_info from task to
>   *_ti_thread_flag().
> 
> Changes since v2:
> - Move membarrier_arch_sched_in() call to finish_task_switch().
> - Check for NULL t->mm in membarrier_arch_fork().
> - Use membarrier_sched_in() in generic code, which invokes the
>   arch-specific membarrier_arch_sched_in(). This fixes allnoconfig
>   build on PowerPC.
> - Move asm/membarrier.h include under CONFIG_MEMBARRIER, fixing
>   allnoconfig build on PowerPC.
> - Build and runtime tested on PowerPC.
> 
> Changes since v3:
> - Simply rely on copy_mm() to copy the membarrier_private_expedited mm
>   field on fork.
> - powerpc: test thread flag instead of reading
>   membarrier_private_expedited in membarrier_arch_fork().
> - powerpc: skip memory barrier in membarrier_arch_sched_in() if coming
>   from kernel thread, since mmdrop() implies a full barrier.
> - Set membarrier_private_expedited to 1 only after arch registration
>   code, thus eliminating a race where concurrent commands could succeed
>   when they should fail if issued concurrently with process
>   registration.
> - Use READ_ONCE() for membarrier_private_expedited field access in
>   membarrier_private_expedited. Matches WRITE_ONCE() performed in
>   process registration.
> 
> Changes since v4:
> - Move powerpc hook from sched_in() to switch_mm(), based on feedback
>   from Nicholas Piggin.

For now, the powerpc approach is okay by me. I plan to test
others (e.g., taking runqueue locks) on larger systems, but that can
be sent as an incremental patch at a later time.

The main thing I would like is for people to review the userspace API.


> diff --git a/include/linux/sched/mm.h b/include/linux/sched/mm.h
> index 3a19c253bdb1..4af1b719c65f 100644
> --- a/include/linux/sched/mm.h
> +++ b/include/linux/sched/mm.h
> @@ -205,4 +205,54 @@ static inline void memalloc_noreclaim_restore(unsigned int flags)
>  	current->flags = (current->flags & ~PF_MEMALLOC) | flags;
>  }
>  
> +#ifdef CONFIG_MEMBARRIER
> +
> +#ifdef CONFIG_ARCH_HAS_MEMBARRIER_HOOKS
> +#include <asm/membarrier.h>
> +#else
> +static inline void membarrier_arch_switch_mm(struct mm_struct *prev,
> +		struct mm_struct *next, struct task_struct *tsk)
> +{
> +}

This is no longer required in architecture independent code, is it?