lkml.org 
[lkml]   [2022]   [Oct]   [18]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
/
Date
SubjectRe: [PATCH v3 4/5] locking/rwsem: Enable direct rwsem lock handoff
From
On 10/18/22 10:13, Mukesh Ojha wrote:
> Hi,
>
> On 10/18/2022 4:44 PM, Hillf Danton wrote:
>> On 17 Oct 2022 17:13:55 -0400 Waiman Long <longman@redhat.com>
>>> @@ -1067,13 +1119,33 @@ rwsem_down_read_slowpath(struct rw_semaphore
>>> *sem, long count, unsigned int stat
>>>               return sem;
>>>           }
>>>           adjustment += RWSEM_FLAG_WAITERS;
>>> +    } else if ((count & RWSEM_FLAG_HANDOFF) &&
>>> +          ((count & RWSEM_LOCK_MASK) == RWSEM_READER_BIAS)) {
>>
>> Could a couple of CPUs go read slow path in parallel?
This is under wait_lock protection. So no parallel execution is possible.
>>
>>> +        /*
>>> +         * If the waiter to be handed off is a reader, this reader
>>> +         * can piggyback on top of top of that.
>>> +         */
>>> +        if (rwsem_first_waiter(sem)->type == RWSEM_WAITING_FOR_READ)
>>> +            adjustment = 0;
>>> +        rwsem_handoff(sem, adjustment, &wake_q);
>>> +
>>> +        if (!adjustment) {
>>> +            raw_spin_unlock_irq(&sem->wait_lock);
>>> +            wake_up_q(&wake_q);
>>> +            return sem;
>>> +        }
>>> +        adjustment = 0;
>>>       }
>>>       rwsem_add_waiter(sem, &waiter);
>>
>> Why can this acquirer pigyback without becoming a waiter?
The idea is to have as much reader parallelism as possible without
writer starvation. In other word, a continuous stream of readers is not
allowed to block out writer. However, there are places where allow one
more reader to get the lock won't cause writer starvation.
>>
>>>   -    /* we're now waiting on the lock, but no longer actively
>>> locking */
>>> -    count = atomic_long_add_return(adjustment, &sem->count);
>>> -
>>> -    rwsem_cond_wake_waiter(sem, count, &wake_q);
>>> +    if (adjustment) {
>>> +        /*
>>> +         * We are now waiting on the lock with no handoff, but no
>>> +         * longer actively locking.
>>> +         */
>>> +        count = atomic_long_add_return(adjustment, &sem->count);
>>> +        rwsem_cond_wake_waiter(sem, count, &wake_q);
>>> +    }
>>>       raw_spin_unlock_irq(&sem->wait_lock);
>>>         if (!wake_q_empty(&wake_q))
>>> @@ -1120,7 +1192,6 @@ static struct rw_semaphore __sched *
>>>   rwsem_down_write_slowpath(struct rw_semaphore *sem, int state)
>>>   {
>>>       struct rwsem_waiter waiter;
>>> -    int null_owner_retries;
>>
>> This reverts 2/5 and feel free to drop it directly.
>
> I think, he intents to tag the first two patches to go to stable
> branches.

This patch is too disruptive to go to the stable branches. Yes, I do
intend the first 2 patches to go into stable branches.

Cheers,
Longman

\
 
 \ /
  Last update: 2022-10-18 19:37    [W:1.183 / U:1.544 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site