Re: [PATCH 6/8] locking/mcs: Better differentiate between mcs variants

[Ingo Molnar <mingo@kernel.org>]

* Davidlohr Bueso <dave@stgolabs.net> wrote:

> We have two flavors of the MCS spinlock: standard and cancelable (osq).
> While each one is independent of the other, we currently mix and match
> them. This patch:
> 
> o Moves osq code out of mcs_spinlock.h (which only deals with the traditional
> version) into include/linux/osq_lock.h. No unnecessary code is added to the
> more global header file, anything locks that make use of osq must include
> it anyway.
> 
> o Renames mcs_spinlock.c to osq_lock.c. This file only contains osq code.
> 
> o Introduces a CONFIG_LOCK_SPIN_ON_OWNER in order to only build osq_lock
> if there is support for it.
> 
> Signed-off-by: Davidlohr Bueso <dbueso@suse.de>
> ---
>  include/linux/osq_lock.h      |  12 ++-
>  kernel/Kconfig.locks          |   4 +
>  kernel/locking/Makefile       |   3 +-
>  kernel/locking/mcs_spinlock.c | 208 ------------------------------------------
>  kernel/locking/mcs_spinlock.h |  16 ----
>  kernel/locking/osq_lock.c     | 203 +++++++++++++++++++++++++++++++++++++++++
>  6 files changed, 219 insertions(+), 227 deletions(-)
>  delete mode 100644 kernel/locking/mcs_spinlock.c
>  create mode 100644 kernel/locking/osq_lock.c

x86 defconfig produces:

  kernel/locking/mcs_spinlock.h:81:2: error: implicit declaration of function ‘ASSIGN_ONCE’ [-Werror=implicit-function-declaration]

Also, this patch should really be split into three separate (and 
tested) patches.

Thanks,

	Ingo