From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1758474AbeD0P4J (ORCPT ); Fri, 27 Apr 2018 11:56:09 -0400 Received: from mx0b-001b2d01.pphosted.com ([148.163.158.5]:52412 "EHLO mx0a-001b2d01.pphosted.com" rhost-flags-OK-OK-OK-FAIL) by vger.kernel.org with ESMTP id S1752471AbeD0P4I (ORCPT ); Fri, 27 Apr 2018 11:56:08 -0400 Date: Fri, 27 Apr 2018 08:57:01 -0700 From: "Paul E. McKenney" To: Joel Fernandes Cc: linux-kernel@vger.kernel.org, Steven Rostedt , Peter Zilstra , Ingo Molnar , Mathieu Desnoyers , Tom Zanussi , Namhyung Kim , Thomas Glexiner , Boqun Feng , Frederic Weisbecker , Randy Dunlap , Masami Hiramatsu , Fenguang Wu , Baohong Liu , Vedang Patel , kernel-team@android.com Subject: Re: [PATCH RFC] tracepoint: Introduce tracepoint callbacks executing with preempt on Reply-To: paulmck@linux.vnet.ibm.com References: <20180427042656.190746-1-joelaf@google.com> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <20180427042656.190746-1-joelaf@google.com> User-Agent: Mutt/1.5.21 (2010-09-15) X-TM-AS-GCONF: 00 x-cbid: 18042715-0008-0000-0000-000002FF3F6C X-IBM-SpamModules-Scores: X-IBM-SpamModules-Versions: BY=3.00008932; HX=3.00000241; KW=3.00000007; PH=3.00000004; SC=3.00000257; SDB=6.01024075; UDB=6.00522844; IPR=6.00803305; MB=3.00020808; MTD=3.00000008; XFM=3.00000015; UTC=2018-04-27 15:55:49 X-IBM-AV-DETECTION: SAVI=unused REMOTE=unused XFE=unused x-cbparentid: 18042715-0009-0000-0000-0000390D33B7 Message-Id: <20180427155701.GL26088@linux.vnet.ibm.com> X-Proofpoint-Virus-Version: vendor=fsecure engine=2.50.10434:,, definitions=2018-04-27_04:,, signatures=0 X-Proofpoint-Spam-Details: rule=outbound_notspam policy=outbound score=0 priorityscore=1501 malwarescore=0 suspectscore=1 phishscore=0 bulkscore=0 spamscore=0 clxscore=1015 lowpriorityscore=0 impostorscore=0 adultscore=0 classifier=spam adjust=0 reason=mlx scancount=1 engine=8.0.1-1709140000 definitions=main-1804270153 Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org On Thu, Apr 26, 2018 at 09:26:56PM -0700, Joel Fernandes wrote: > In recent tests with IRQ on/off tracepoints, a large performance > overhead ~10% is noticed when running hackbench. This is root caused to > calls to rcu_irq_enter_irqson and rcu_irq_exit_irqson from the > tracepoint code. Following a long discussion on the list [1] about this, > we concluded that srcu is a better alternative for use during rcu idle. > Although it does involve extra barriers, its lighter than the sched-rcu > version which has to do additional RCU calls to notify RCU idle about > entry into RCU sections. > > In this patch, we change the underlying implementation of the > trace_*_rcuidle API to use SRCU. This has shown to improve performance > alot for the high frequency irq enable/disable tracepoints. > > In the future, we can add a new may_sleep API which can use this > infrastructure for callbacks that actually can sleep which will support > Mathieu's usecase of blocking probes. > > Test: Tested idle and preempt/irq tracepoints. Looks good overall! One question and a few comments below. Thanx, Paul > [1] https://patchwork.kernel.org/patch/10344297/ > > Cc: Steven Rostedt > Cc: Peter Zilstra > Cc: Ingo Molnar > Cc: Mathieu Desnoyers > Cc: Tom Zanussi > Cc: Namhyung Kim > Cc: Thomas Glexiner > Cc: Boqun Feng > Cc: Paul McKenney > Cc: Frederic Weisbecker > Cc: Randy Dunlap > Cc: Masami Hiramatsu > Cc: Fenguang Wu > Cc: Baohong Liu > Cc: Vedang Patel > Cc: kernel-team@android.com > Signed-off-by: Joel Fernandes > --- > include/linux/tracepoint.h | 37 +++++++++++++++++++++++++++++-------- > kernel/tracepoint.c | 10 +++++++++- > 2 files changed, 38 insertions(+), 9 deletions(-) > > diff --git a/include/linux/tracepoint.h b/include/linux/tracepoint.h > index c94f466d57ef..a1c1987de423 100644 > --- a/include/linux/tracepoint.h > +++ b/include/linux/tracepoint.h > @@ -15,6 +15,7 @@ > */ > > #include > +#include > #include > #include > #include > @@ -33,6 +34,8 @@ struct trace_eval_map { > > #define TRACEPOINT_DEFAULT_PRIO 10 > > +extern struct srcu_struct tracepoint_srcu; > + > extern int > tracepoint_probe_register(struct tracepoint *tp, void *probe, void *data); > extern int > @@ -77,6 +80,7 @@ int unregister_tracepoint_module_notifier(struct notifier_block *nb) > */ > static inline void tracepoint_synchronize_unregister(void) > { > + synchronize_srcu(&tracepoint_srcu); > synchronize_sched(); > } > > @@ -129,18 +133,26 @@ extern void syscall_unregfunc(void); > * as "(void *, void)". The DECLARE_TRACE_NOARGS() will pass in just > * "void *data", where as the DECLARE_TRACE() will pass in "void *data, proto". > */ > -#define __DO_TRACE(tp, proto, args, cond, rcucheck) \ > +#define __DO_TRACE(tp, proto, args, cond, preempt_on) \ > do { \ > struct tracepoint_func *it_func_ptr; \ > void *it_func; \ > void *__data; \ > + int __maybe_unused idx = 0; \ > \ > if (!(cond)) \ > return; \ > - if (rcucheck) \ > - rcu_irq_enter_irqson(); \ > - rcu_read_lock_sched_notrace(); \ > - it_func_ptr = rcu_dereference_sched((tp)->funcs); \ > + if (preempt_on) { \ > + WARN_ON_ONCE(in_nmi()); /* no srcu from nmi */ \ Very good on this check, thank you! > + idx = srcu_read_lock(&tracepoint_srcu); \ Hmmm... Do I need to create a _notrace variant of srcu_read_lock() and srcu_read_unlock()? > + it_func_ptr = srcu_dereference((tp)->funcs, \ > + &tracepoint_srcu); \ > + } else { \ > + rcu_read_lock_sched_notrace(); \ > + it_func_ptr = \ > + rcu_dereference_sched((tp)->funcs); \ > + } \ > + \ > if (it_func_ptr) { \ > do { \ > it_func = (it_func_ptr)->func; \ > @@ -148,12 +160,21 @@ extern void syscall_unregfunc(void); > ((void(*)(proto))(it_func))(args); \ > } while ((++it_func_ptr)->func); \ > } \ > - rcu_read_unlock_sched_notrace(); \ > - if (rcucheck) \ > - rcu_irq_exit_irqson(); \ > + \ > + if (preempt_on) \ > + srcu_read_unlock(&tracepoint_srcu, idx); \ > + else \ > + rcu_read_unlock_sched_notrace(); \ > } while (0) > > #ifndef MODULE > +/* > + * This is for tracepoints that may be called from an rcu idle path. To make > + * sure we its safe in the idle path, use the srcu instead of sched-rcu by > + * specifying the preempt_on flag below. This has the obvious effect that any > + * callback expecting preemption to be disabled should explicitly do so since > + * with srcu it'd run with preempt on. > + */ > #define __DECLARE_TRACE_RCU(name, proto, args, cond, data_proto, data_args) \ > static inline void trace_##name##_rcuidle(proto) \ > { \ > diff --git a/kernel/tracepoint.c b/kernel/tracepoint.c > index 671b13457387..b3b1d65a2460 100644 > --- a/kernel/tracepoint.c > +++ b/kernel/tracepoint.c > @@ -31,6 +31,9 @@ > extern struct tracepoint * const __start___tracepoints_ptrs[]; > extern struct tracepoint * const __stop___tracepoints_ptrs[]; > > +DEFINE_SRCU(tracepoint_srcu); > +EXPORT_SYMBOL_GPL(tracepoint_srcu); > + > /* Set to 1 to enable tracepoint debug output */ > static const int tracepoint_debug; > > @@ -67,11 +70,16 @@ static inline void *allocate_probes(int count) > return p == NULL ? NULL : p->probes; > } > > -static void rcu_free_old_probes(struct rcu_head *head) > +static void srcu_free_old_probes(struct rcu_head *head) > { > kfree(container_of(head, struct tp_probes, rcu)); > } > > +static void rcu_free_old_probes(struct rcu_head *head) > +{ > + call_srcu(&tracepoint_srcu, head, srcu_free_old_probes); Chaining them through the two callbacks works, good! > +} > + > static inline void release_probes(struct tracepoint_func *old) > { > if (old) { > -- > 2.17.0.441.gb46fe60e1d-goog >