From: Waiman Long <waiman.long@hp.com>
To: Peter Zijlstra <peterz@infradead.org>
Cc: Thomas Gleixner <tglx@linutronix.de>,
Ingo Molnar <mingo@redhat.com>, "H. Peter Anvin" <hpa@zytor.com>,
Arnd Bergmann <arnd@arndb.de>,
linux-arch@vger.kernel.org, x86@kernel.org,
linux-kernel@vger.kernel.org,
Steven Rostedt <rostedt@goodmis.org>,
Andrew Morton <akpm@linux-foundation.org>,
Michel Lespinasse <walken@google.com>,
Andi Kleen <andi@firstfloor.org>, Rik van Riel <riel@redhat.com>,
"Paul E. McKenney" <paulmck@linux.vnet.ibm.com>,
Linus Torvalds <torvalds@linux-foundation.org>,
Raghavendra K T <raghavendra.kt@linux.vnet.ibm.com>,
George Spelvin <linux@horizon.com>,
Tim Chen <tim.c.chen@linux.intel.com>, "" <aswin@hp.com>,
Scott J Norton <scott.norton@hp.com>
Subject: Re: [PATCH v11 1/4] qrwlock: A queue read/write lock implementation
Date: Fri, 24 Jan 2014 23:39:10 -0500 [thread overview]
Message-ID: <52E33FEE.50805@hp.com> (raw)
In-Reply-To: <20140124082514.GB31570@twins.programming.kicks-ass.net>
On 01/24/2014 03:25 AM, Peter Zijlstra wrote:
> On Thu, Jan 23, 2014 at 11:28:48PM -0500, Waiman Long wrote:
>> +/**
>> + * queue_read_trylock - try to acquire read lock of a queue rwlock
>> + * @lock : Pointer to queue rwlock structure
>> + * Return: 1 if lock acquired, 0 if failed
>> + */
>> +static inline int queue_read_trylock(struct qrwlock *lock)
>> +{
>> + union qrwcnts cnts;
>> +
>> + cnts.rwc = ACCESS_ONCE(lock->cnts.rwc);
>> + if (likely(!cnts.writer)) {
>> + cnts.rwc = (u32)atomic_add_return(_QR_BIAS,&lock->cnts.rwa);
>> + if (likely(!cnts.writer)) {
>> + smp_mb__after_atomic_inc();
> That's superfluous, as atomic_add_return() is documented as being a full
> barrier.
Yes, you are right. I have reviewed the memory_barrier.txt again and
atomic_add_return() is supposed to act as a memory barrier. So no extra
barrier. I will correct that in the next version.
>> + return 1;
>> + }
>> + atomic_sub(_QR_BIAS,&lock->cnts.rwa);
>> + }
>> + return 0;
>> +}
>> +
>> +/**
>> + * queue_write_trylock - try to acquire write lock of a queue rwlock
>> + * @lock : Pointer to queue rwlock structure
>> + * Return: 1 if lock acquired, 0 if failed
>> + */
>> +static inline int queue_write_trylock(struct qrwlock *lock)
>> +{
>> + union qrwcnts old, new;
>> +
>> + old.rwc = ACCESS_ONCE(lock->cnts.rwc);
>> + if (likely(!old.rwc)) {
>> + new.rwc = old.rwc;
>> + new.writer = _QW_LOCKED;
>> + if (likely(cmpxchg(&lock->cnts.rwc, old.rwc, new.rwc)
>> + == old.rwc))
> One could actually use atomic_cmpxchg() and avoid one (ab)use of that
> union :-)
I think either one is fine. I would like to keep the original code if it
is not really a problem.
>> + return 1;
>> + }
>> + return 0;
>> +}
>> +/**
>> + * queue_read_lock - acquire read lock of a queue rwlock
>> + * @lock: Pointer to queue rwlock structure
>> + */
>> +static inline void queue_read_lock(struct qrwlock *lock)
>> +{
>> + union qrwcnts cnts;
>> +
>> + cnts.rwc = atomic_add_return(_QR_BIAS,&lock->cnts.rwa);
>> + if (likely(!cnts.writer)) {
>> + smp_mb__after_atomic_inc();
> Superfluous again.
Will remove that.
>> + return;
>> + queue_write_lock_slowpath(lock);
>> +}
>> +
>> +/**
>> + * queue_read_unlock - release read lock of a queue rwlock
>> + * @lock : Pointer to queue rwlock structure
>> + */
>> +static inline void queue_read_unlock(struct qrwlock *lock)
>> +{
>> + /*
>> + * Atomically decrement the reader count
>> + */
>> + smp_mb__before_atomic_dec();
>> + atomic_sub(_QR_BIAS,&lock->cnts.rwa);
>> +}
>> +
>> +/**
>> + * queue_write_unlock - release write lock of a queue rwlock
>> + * @lock : Pointer to queue rwlock structure
>> + */
>> +static inline void queue_write_unlock(struct qrwlock *lock)
>> +{
>> + /*
>> + * If the writer field is atomic, it can be cleared directly.
>> + * Otherwise, an atomic subtraction will be used to clear it.
>> + */
>> + if (__native_word(lock->cnts.writer))
>> + smp_store_release(&lock->cnts.writer, 0);
>> + else {
>> + smp_mb__before_atomic_dec();
>> + atomic_sub(_QW_LOCKED,&lock->cnts.rwa);
>> + }
> Missing {}, Documentation/CodingStyle Chapter 3 near the very end.
Thank for spotting that. Will fix it in the next version.
-Longman
next prev parent reply other threads:[~2014-01-25 4:39 UTC|newest]
Thread overview: 33+ messages / expand[flat|nested] mbox.gz Atom feed top
2014-01-24 4:28 [PATCH v11 0/4] Introducing a queue read/write lock implementation Waiman Long
2014-01-24 4:28 ` [PATCH v11 1/4] qrwlock: A " Waiman Long
2014-01-24 8:25 ` Peter Zijlstra
2014-01-25 4:39 ` Waiman Long [this message]
2014-01-24 4:28 ` [PATCH v11 2/4] qrwlock, x86: Enable x86 to use queue read/write lock Waiman Long
2014-01-24 4:28 ` [PATCH v11 3/4] qrwlock, x86: Add char and short as atomic data type in x86 Waiman Long
2014-01-24 4:28 ` [PATCH v11 4/4] qrwlock: Use the mcs_spinlock helper functions for MCS queuing Waiman Long
2014-01-24 8:26 ` Peter Zijlstra
2014-01-25 4:30 ` Waiman Long
2014-01-30 13:04 ` [PATCH v11 0/4] Introducing a queue read/write lock implementation Peter Zijlstra
2014-01-30 15:17 ` Peter Zijlstra
2014-01-30 15:43 ` Waiman Long
2014-01-30 15:50 ` Waiman Long
2014-01-30 15:53 ` Peter Zijlstra
2014-01-30 15:44 ` Peter Zijlstra
2014-01-30 17:52 ` Will Deacon
2014-01-30 18:05 ` Peter Zijlstra
2014-01-30 18:11 ` Will Deacon
2014-01-30 18:16 ` Peter Zijlstra
2014-01-31 9:26 ` Peter Zijlstra
2014-01-31 10:03 ` George Spelvin
2014-01-31 10:17 ` Peter Zijlstra
2014-01-31 11:30 ` Peter Zijlstra
2014-02-01 23:22 ` Paul E. McKenney
2014-01-31 18:59 ` Waiman Long
2014-01-31 19:47 ` Peter Zijlstra
2014-02-01 10:38 ` George Spelvin
2014-01-31 20:14 ` Peter Zijlstra
2014-01-31 21:09 ` Waiman Long
2014-02-01 1:29 ` Davidlohr Bueso
2014-02-02 9:03 ` Ingo Molnar
2014-02-03 11:38 ` Peter Zijlstra
2014-02-06 3:08 ` Waiman Long
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=52E33FEE.50805@hp.com \
--to=waiman.long@hp.com \
--cc=akpm@linux-foundation.org \
--cc=andi@firstfloor.org \
--cc=arnd@arndb.de \
--cc=aswin@hp.com \
--cc=hpa@zytor.com \
--cc=linux-arch@vger.kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=linux@horizon.com \
--cc=mingo@redhat.com \
--cc=paulmck@linux.vnet.ibm.com \
--cc=peterz@infradead.org \
--cc=raghavendra.kt@linux.vnet.ibm.com \
--cc=riel@redhat.com \
--cc=rostedt@goodmis.org \
--cc=scott.norton@hp.com \
--cc=tglx@linutronix.de \
--cc=tim.c.chen@linux.intel.com \
--cc=torvalds@linux-foundation.org \
--cc=walken@google.com \
--cc=x86@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).