From: Waiman Long <longman@redhat.com>
To: Peter Zijlstra <peterz@infradead.org>
Cc: mingo@redhat.com, will@kernel.org, linux-kernel@vger.kernel.org,
boqun.feng@gmail.com
Subject: Re: [PATCH 3/6] locking/rwsem: Rework writer wakeup
Date: Sun, 26 Feb 2023 16:31:15 -0500 [thread overview]
Message-ID: <1147abb3-fb72-dd63-8e32-25ff8000972e@redhat.com> (raw)
In-Reply-To: <Y/tJ2n1e22YhsZ17@hirez.programming.kicks-ass.net>
On 2/26/23 07:00, Peter Zijlstra wrote:
> On Thu, Feb 23, 2023 at 04:38:08PM -0500, Waiman Long wrote:
>
>>> @@ -1143,54 +1138,36 @@ rwsem_down_write_slowpath(struct rw_sema
>>> } else {
>>> atomic_long_or(RWSEM_FLAG_WAITERS, &sem->count);
>>> }
>>> + raw_spin_unlock_irq(&sem->wait_lock);
>>> /* wait until we successfully acquire the lock */
>>> - set_current_state(state);
>>> trace_contention_begin(sem, LCB_F_WRITE);
>>> for (;;) {
>>> - if (rwsem_try_write_lock(sem, &waiter)) {
>>> - /* rwsem_try_write_lock() implies ACQUIRE on success */
>>> + set_current_state(state);
>>> + if (!smp_load_acquire(&waiter.task)) {
>>> + /* Matches rwsem_waiter_wake()'s smp_store_release(). */
>>> break;
>>> }
>>> -
>>> - raw_spin_unlock_irq(&sem->wait_lock);
>>> -
>>> - if (signal_pending_state(state, current))
>>> - goto out_nolock;
>>> -
>>> - /*
>>> - * After setting the handoff bit and failing to acquire
>>> - * the lock, attempt to spin on owner to accelerate lock
>>> - * transfer. If the previous owner is a on-cpu writer and it
>>> - * has just released the lock, OWNER_NULL will be returned.
>>> - * In this case, we attempt to acquire the lock again
>>> - * without sleeping.
>>> - */
>>> - if (waiter.handoff_set) {
>>> - enum owner_state owner_state;
>>> -
>>> - owner_state = rwsem_spin_on_owner(sem);
>>> - if (owner_state == OWNER_NULL)
>>> - goto trylock_again;
>>> + if (signal_pending_state(state, current)) {
>>> + raw_spin_lock_irq(&sem->wait_lock);
>>> + if (waiter.task)
>>> + goto out_nolock;
>>> + raw_spin_unlock_irq(&sem->wait_lock);
>>> + /* Ordered by sem->wait_lock against rwsem_mark_wake(). */
>>> + break;
>>> }
>>> -
>>> schedule_preempt_disabled();
>>> lockevent_inc(rwsem_sleep_writer);
>>> - set_current_state(state);
>>> -trylock_again:
>>> - raw_spin_lock_irq(&sem->wait_lock);
>>> }
>>> __set_current_state(TASK_RUNNING);
>>> - raw_spin_unlock_irq(&sem->wait_lock);
>>> lockevent_inc(rwsem_wlock);
>>> trace_contention_end(sem, 0);
>>> return sem;
>>> out_nolock:
>>> - __set_current_state(TASK_RUNNING);
>>> - raw_spin_lock_irq(&sem->wait_lock);
>>> rwsem_del_wake_waiter(sem, &waiter, &wake_q);
>>> + __set_current_state(TASK_RUNNING);
>>> lockevent_inc(rwsem_wlock_fail);
>>> trace_contention_end(sem, -EINTR);
>>> return ERR_PTR(-EINTR);
>> I believe it is better to change state inside the wait_lock critical section
>> to provide a release barrier for free.
> I can't follow... a release for what? Note that the reader slowpath has
> this exact form already.\
You are right. I forgot that we don't need synchronization when setting
state to TASK_RUNNING.
Cheers,
Longman
next prev parent reply other threads:[~2023-02-26 21:32 UTC|newest]
Thread overview: 27+ messages / expand[flat|nested] mbox.gz Atom feed top
2023-02-23 12:26 [PATCH 0/6] locking/rwsem: Rework writer wakeup and handoff Peter Zijlstra
2023-02-23 12:26 ` [PATCH 1/6] locking/rwsem: Minor code refactoring in rwsem_mark_wake() Peter Zijlstra
2023-02-23 12:26 ` [PATCH 2/6] locking/rwsem: Enforce queueing when HANDOFF Peter Zijlstra
2023-02-23 12:26 ` [PATCH 3/6] locking/rwsem: Rework writer wakeup Peter Zijlstra
2023-02-23 21:38 ` Waiman Long
2023-02-26 11:58 ` Peter Zijlstra
2023-02-26 12:00 ` Peter Zijlstra
2023-02-26 21:31 ` Waiman Long [this message]
2023-02-26 11:59 ` Peter Zijlstra
2023-02-26 15:04 ` Peter Zijlstra
2023-02-26 16:51 ` Peter Zijlstra
2023-02-27 0:22 ` Waiman Long
2023-02-27 10:31 ` Peter Zijlstra
2023-02-27 20:16 ` Waiman Long
2023-03-20 8:12 ` Peter Zijlstra
2023-03-20 17:36 ` Waiman Long
2023-02-23 12:26 ` [PATCH 4/6] locking/rwsem: Split out rwsem_reader_wake() Peter Zijlstra
2023-02-23 12:26 ` [PATCH 5/6] locking/rwsem: Unify wait loop Peter Zijlstra
2023-02-23 19:31 ` Boqun Feng
2023-02-24 1:33 ` Boqun Feng
2023-02-26 12:01 ` Peter Zijlstra
2023-02-26 18:22 ` Boqun Feng
2023-02-23 22:45 ` Waiman Long
2023-02-26 16:15 ` Peter Zijlstra
2023-02-23 12:26 ` [PATCH 6/6] locking/rwsem: Use the force Peter Zijlstra
2023-02-24 1:19 ` [PATCH 0/6] locking/rwsem: Rework writer wakeup and handoff Waiman Long
2023-02-24 11:55 ` Jiri Wiesner
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=1147abb3-fb72-dd63-8e32-25ff8000972e@redhat.com \
--to=longman@redhat.com \
--cc=boqun.feng@gmail.com \
--cc=linux-kernel@vger.kernel.org \
--cc=mingo@redhat.com \
--cc=peterz@infradead.org \
--cc=will@kernel.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).