lkml.org 
[lkml]   [2022]   [Jun]   [28]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
/
Date
SubjectRe: [PATCH V7 4/5] asm-generic: spinlock: Add combo spinlock (ticket & queued)
From
On 6/28/22 21:17, Guo Ren wrote:
> On Wed, Jun 29, 2022 at 2:13 AM Waiman Long <longman@redhat.com> wrote:
>> On 6/28/22 04:17, guoren@kernel.org wrote:
>>> From: Guo Ren <guoren@linux.alibaba.com>
>>>
>>> Some architecture has a flexible requirement on the type of spinlock.
>>> Some LL/SC architectures of ISA don't force micro-arch to give a strong
>>> forward guarantee. Thus different kinds of memory model micro-arch would
>>> come out in one ISA. The ticket lock is suitable for exclusive monitor
>>> designed LL/SC micro-arch with limited cores and "!NUMA". The
>>> queue-spinlock could deal with NUMA/large-scale scenarios with a strong
>>> forward guarantee designed LL/SC micro-arch.
>>>
>>> So, make the spinlock a combo with feature.
>>>
>>> Signed-off-by: Guo Ren <guoren@linux.alibaba.com>
>>> Signed-off-by: Guo Ren <guoren@kernel.org>
>>> Cc: Peter Zijlstra (Intel) <peterz@infradead.org>
>>> Cc: Arnd Bergmann <arnd@arndb.de>
>>> Cc: Palmer Dabbelt <palmer@rivosinc.com>
>>> ---
>>> include/asm-generic/spinlock.h | 43 ++++++++++++++++++++++++++++++++--
>>> kernel/locking/qspinlock.c | 2 ++
>>> 2 files changed, 43 insertions(+), 2 deletions(-)
>>>
>>> diff --git a/include/asm-generic/spinlock.h b/include/asm-generic/spinlock.h
>>> index f41dc7c2b900..a9b43089bf99 100644
>>> --- a/include/asm-generic/spinlock.h
>>> +++ b/include/asm-generic/spinlock.h
>>> @@ -28,34 +28,73 @@
>>> #define __ASM_GENERIC_SPINLOCK_H
>>>
>>> #include <asm-generic/ticket_spinlock.h>
>>> +#ifdef CONFIG_ARCH_USE_QUEUED_SPINLOCKS
>>> +#include <linux/jump_label.h>
>>> +#include <asm-generic/qspinlock.h>
>>> +
>>> +DECLARE_STATIC_KEY_TRUE(use_qspinlock_key);
>>> +#endif
>>> +
>>> +#undef arch_spin_is_locked
>>> +#undef arch_spin_is_contended
>>> +#undef arch_spin_value_unlocked
>>> +#undef arch_spin_lock
>>> +#undef arch_spin_trylock
>>> +#undef arch_spin_unlock
>>>
>>> static __always_inline void arch_spin_lock(arch_spinlock_t *lock)
>>> {
>>> - ticket_spin_lock(lock);
>>> +#ifdef CONFIG_ARCH_USE_QUEUED_SPINLOCKS
>>> + if (static_branch_likely(&use_qspinlock_key))
>>> + queued_spin_lock(lock);
>>> + else
>>> +#endif
>>> + ticket_spin_lock(lock);
>>> }
>> Why do you use a static key to control whether to use qspinlock or
>> ticket lock? In the next patch, you have
>>
>> +#if !defined(CONFIG_NUMA) && defined(CONFIG_QUEUED_SPINLOCKS)
>> + static_branch_disable(&use_qspinlock_key);
>> +#endif
>>
>> So the current config setting determines if qspinlock will be used, not
>> some boot time parameter that user needs to specify. This patch will
>> just add useless code to lock/unlock sites. I don't see any benefit of
>> doing that.
> This is a startup patch for riscv. next, we could let vendors make choices.
> I'm not sure they like cmdline or vendor-specific errata style.
>
> Eventually, we would let one riscv Image support all machines, some
> use ticket-lock, and some use qspinlock.

OK. Maybe you can postpone this combo spinlock until there is a good use
case for it. Upstream usually don't accept patches that have no good use
case yet.

Cheers,
Longman

\
 
 \ /
  Last update: 2022-06-29 03:34    [W:0.066 / U:0.068 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site