From mboxrd@z Thu Jan 1 00:00:00 1970 From: Mathieu Desnoyers Subject: Re: [PATCHv5 2/2] memory barrier: adding smp_mb__after_lock Date: Tue, 7 Jul 2009 10:01:35 -0400 Message-ID: <20090707140135.GA5506@Krystal> References: <20090703081219.GE2902@jolsa.lab.eng.brq.redhat.com> <20090703081445.GG2902@jolsa.lab.eng.brq.redhat.com> <20090703090606.GA3902@elte.hu> <4A4DCD54.1080908@gmail.com> <20090703092438.GE3902@elte.hu> <20090703095659.GA4518@jolsa.lab.eng.brq.redhat.com> <20090703102530.GD32128@elte.hu> <20090703111848.GA10267@jolsa.lab.eng.brq.redhat.com> <20090707101816.GA6619@jolsa.lab.eng.brq.redhat.com> <20090707134601.GB6619@jolsa.lab.eng.brq.redhat.com> Mime-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: QUOTED-PRINTABLE Cc: Ingo Molnar , Eric Dumazet , Peter Zijlstra , netdev@vger.kernel.org, linux-kernel@vger.kernel.org, fbl@redhat.com, nhorman@redhat.com, davem@redhat.com, htejun@gmail.com, jarkao2@gmail.com, oleg@redhat.com, davidel@xmailserver.org To: Jiri Olsa Return-path: Received: from tomts43.bellnexxia.net ([209.226.175.110]:41921 "EHLO tomts43-srv.bellnexxia.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1755203AbZGGOCj convert rfc822-to-8bit (ORCPT ); Tue, 7 Jul 2009 10:02:39 -0400 Content-Disposition: inline In-Reply-To: <20090707134601.GB6619@jolsa.lab.eng.brq.redhat.com> Sender: netdev-owner@vger.kernel.org List-ID: * Jiri Olsa (jolsa@redhat.com) wrote: > On Tue, Jul 07, 2009 at 12:18:16PM +0200, Jiri Olsa wrote: > > On Fri, Jul 03, 2009 at 01:18:48PM +0200, Jiri Olsa wrote: > > > On Fri, Jul 03, 2009 at 12:25:30PM +0200, Ingo Molnar wrote: > > > >=20 > > > > * Jiri Olsa wrote: > > > >=20 > > > > > On Fri, Jul 03, 2009 at 11:24:38AM +0200, Ingo Molnar wrote: > > > > > >=20 > > > > > > * Eric Dumazet wrote: > > > > > >=20 > > > > > > > Ingo Molnar a =E9crit : > > > > > > > > * Jiri Olsa wrote: > > > > > > > >=20 > > > > > > > >> +++ b/arch/x86/include/asm/spinlock.h > > > > > > > >> @@ -302,4 +302,7 @@ static inline void __raw_write_unl= ock(raw_rwlock_t *rw) > > > > > > > >> #define _raw_read_relax(lock) cpu_relax() > > > > > > > >> #define _raw_write_relax(lock) cpu_relax() > > > > > > > >> =20 > > > > > > > >> +/* The {read|write|spin}_lock() on x86 are full memor= y barriers. */ > > > > > > > >> +#define smp_mb__after_lock() do { } while (0) > > > > > > > >=20 > > > > > > > > Two small stylistic comments, please make this an inlin= e function: > > > > > > > >=20 > > > > > > > > static inline void smp_mb__after_lock(void) { } > > > > > > > > #define smp_mb__after_lock > > > > > > > >=20 > > > > > > > > (untested) > > > > > > > >=20 > > > > > > > >> +/* The lock does not imply full memory barrier. */ > > > > > > > >> +#ifndef smp_mb__after_lock > > > > > > > >> +#define smp_mb__after_lock() smp_mb() > > > > > > > >> +#endif > > > > > > > >=20 > > > > > > > > ditto. > > > > > > > >=20 > > > > > > > > Ingo > > > > > > >=20 > > > > > > > This was following existing implementations of various sm= p_mb__??? helpers : > > > > > > >=20 > > > > > > > # grep -4 smp_mb__before_clear_bit include/asm-generic/bi= tops.h > > > > > > >=20 > > > > > > > /* > > > > > > > * clear_bit may not imply a memory barrier > > > > > > > */ > > > > > > > #ifndef smp_mb__before_clear_bit > > > > > > > #define smp_mb__before_clear_bit() smp_mb() > > > > > > > #define smp_mb__after_clear_bit() smp_mb() > > > > > > > #endif > > > > > >=20 > > > > > > Did i mention that those should be fixed too? :-) > > > > > >=20 > > > > > > Ingo > > > > >=20 > > > > > ok, could I include it in the 2/2 or you prefer separate patc= h? > > > >=20 > > > > depends on whether it will regress ;-) > > > >=20 > > > > If it regresses, it's better to have it separate. If it wont, i= t can=20 > > > > be included. If unsure, default to the more conservative option= =2E > > > >=20 > > > > Ingo > > >=20 > > >=20 > > > how about this..=20 > > > and similar change for smp_mb__before_clear_bit in a separate pat= ch > > >=20 > > >=20 > > > diff --git a/arch/x86/include/asm/spinlock.h b/arch/x86/include/a= sm/spinlock.h > > > index b7e5db8..4e77853 100644 > > > --- a/arch/x86/include/asm/spinlock.h > > > +++ b/arch/x86/include/asm/spinlock.h > > > @@ -302,4 +302,8 @@ static inline void __raw_write_unlock(raw_rwl= ock_t *rw) > > > #define _raw_read_relax(lock) cpu_relax() > > > #define _raw_write_relax(lock) cpu_relax() > > > =20 > > > +/* The {read|write|spin}_lock() on x86 are full memory barriers.= */ > > > +static inline void smp_mb__after_lock(void) { } > > > +#define ARCH_HAS_SMP_MB_AFTER_LOCK > > > + > > > #endif /* _ASM_X86_SPINLOCK_H */ > > > diff --git a/include/linux/spinlock.h b/include/linux/spinlock.h > > > index 252b245..4be57ab 100644 > > > --- a/include/linux/spinlock.h > > > +++ b/include/linux/spinlock.h > > > @@ -132,6 +132,11 @@ do { \ > > > #endif /*__raw_spin_is_contended*/ > > > #endif > > > =20 > > > +/* The lock does not imply full memory barrier. */ > > > +#ifndef ARCH_HAS_SMP_MB_AFTER_LOCK > > > +static inline void smp_mb__after_lock(void) { smp_mb(); } > > > +#endif > > > + > > > /** > > > * spin_unlock_wait - wait until the spinlock gets unlocked > > > * @lock: the spinlock in question. > > > diff --git a/include/net/sock.h b/include/net/sock.h > > > index 4eb8409..98afcd9 100644 > > > --- a/include/net/sock.h > > > +++ b/include/net/sock.h > > > @@ -1271,6 +1271,9 @@ static inline int sk_has_allocations(const = struct sock *sk) > > > * in its cache, and so does the tp->rcv_nxt update on CPU2 side= =2E The CPU1 > > > * could then endup calling schedule and sleep forever if there = are no more > > > * data on the socket. > > > + * > > > + * The sk_has_helper is always called right after a call to read= _lock, so we > > > + * can use smp_mb__after_lock barrier. > > > */ > > > static inline int sk_has_sleeper(struct sock *sk) > > > { > > > @@ -1280,7 +1283,7 @@ static inline int sk_has_sleeper(struct soc= k *sk) > > > * > > > * This memory barrier is paired in the sock_poll_wait. > > > */ > > > - smp_mb(); > > > + smp_mb__after_lock(); > > > return sk->sk_sleep && waitqueue_active(sk->sk_sleep); > > > } > > > =20 > >=20 > > any feedback on this?=20 > > I'd send v6 if this way is acceptable.. > >=20 > > thanks, > > jirka >=20 > also I checked the smp_mb__before_clear_bit/smp_mb__after_clear_bit a= nd > it is used quite extensivelly. >=20 > I'd prefer to send it in a separate patch, so we can move on with the= =20 > changes I've sent so far.. >=20 As with any optimization (and this is one that adds a semantic that wil= l just grow the memory barrier/locking rule complexity), it should come with performance benchmarks showing real-life improvements. Otherwise I'd recommend sticking to smp_mb() if this execution path is not that critical, or to move to RCU if it's _that_ critical. A valid argument would be if the data structures protected are so complex that RCU is out of question but still the few cycles saved by removing a memory barrier are really significant. And even then, the proper solution would be more something like a __read_lock()+smp_mb+smp_mb+__read_unlock(), so we get the performance improvements on architectures other than x86 as well. So in all cases, I don't think the smp_mb__after_lock() is the appropriate solution. Mathieu > regards, > jirka --=20 Mathieu Desnoyers OpenPGP key fingerprint: 8CD5 52C3 8E3C 4140 715F BA06 3F25 A8FE 3BAE = 9A68