Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1752042Ab3FZAUA (ORCPT ); Tue, 25 Jun 2013 20:20:00 -0400 Received: from mga02.intel.com ([134.134.136.20]:57604 "EHLO mga02.intel.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751385Ab3FZATx (ORCPT ); Tue, 25 Jun 2013 20:19:53 -0400 X-ExtLoop1: 1 X-IronPort-AV: E=Sophos;i="4.87,940,1363158000"; d="scan'208";a="359589793" Subject: Re: Performance regression from switching lock to rw-sem for anon-vma tree From: Tim Chen To: Ingo Molnar Cc: Ingo Molnar , Andrea Arcangeli , Mel Gorman , "Shi, Alex" , Andi Kleen , Andrew Morton , Michel Lespinasse , Davidlohr Bueso , "Wilcox, Matthew R" , Dave Hansen , Peter Zijlstra , Rik van Riel , linux-kernel@vger.kernel.org, linux-mm In-Reply-To: <1371660831.27102.663.camel@schen9-DESK> References: <1371165992.27102.573.camel@schen9-DESK> <20130619131611.GC24957@gmail.com> <1371660831.27102.663.camel@schen9-DESK> Content-Type: text/plain; charset="UTF-8" Date: Tue, 25 Jun 2013 17:19:56 -0700 Message-ID: <1372205996.22432.119.camel@schen9-DESK> Mime-Version: 1.0 X-Mailer: Evolution 2.32.3 (2.32.3-1.fc14) Content-Transfer-Encoding: 7bit Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Length: 5087 Lines: 134 On Wed, 2013-06-19 at 09:53 -0700, Tim Chen wrote: > On Wed, 2013-06-19 at 15:16 +0200, Ingo Molnar wrote: > > > > vmstat for mutex implementation: > > > procs -----------memory---------- ---swap-- -----io---- --system-- -----cpu----- > > > r b swpd free buff cache si so bi bo in cs us sy id wa st > > > 38 0 0 130957920 47860 199956 0 0 0 56 236342 476975 14 72 14 0 0 > > > 41 0 0 130938560 47860 219900 0 0 0 0 236816 479676 14 72 14 0 0 > > > > > > vmstat for rw-sem implementation (3.10-rc4) > > > procs -----------memory---------- ---swap-- -----io---- --system-- -----cpu----- > > > r b swpd free buff cache si so bi bo in cs us sy id wa st > > > 40 0 0 130933984 43232 202584 0 0 0 0 321817 690741 13 71 16 0 0 > > > 39 0 0 130913904 43232 224812 0 0 0 0 322193 692949 13 71 16 0 0 > > > > It appears the main difference is that the rwsem variant context-switches > > about 36% more than the mutex version, right? > > > > I'm wondering how that's possible - the lock is mostly write-locked, > > correct? So the lock-stealing from Davidlohr Bueso and Michel Lespinasse > > ought to have brought roughly the same lock-stealing behavior as mutexes > > do, right? > > > > So the next analytical step would be to figure out why rwsem lock-stealing > > is not behaving in an equivalent fashion on this workload. Do readers come > > in frequently enough to disrupt write-lock-stealing perhaps? Ingo, I did some instrumentation on the write lock failure path. I found that for the exim workload, there are no readers blocking for the rwsem when write locking failed. The lock stealing is successful for 9.1% of the time and the rest of the write lock failure caused the writer to go to sleep. About 1.4% of the writers sleep more than once. Majority of the writers sleep once. It is weird that lock stealing is not successful more often. Tim > > > > Context-switch call-graph profiling might shed some light on where the > > extra context switches come from... > > > > Something like: > > > > perf record -g -e sched:sched_switch --filter 'prev_state != 0' -a sleep 1 > > > > or a variant thereof might do the trick. > > > > Ingo, > > It appears that we are having much more down write failure causing a process to > block vs going to the slow path for the mutex case. > > Here's the profile data from > perf record -g -e sched:sched_switch --filter 'prev_state != 0' -a sleep 1 > > 3.10-rc4 (mutex implementation context switch profile) > > - 59.51% exim [kernel.kallsyms] [k] perf_trace_sched_switch > - perf_trace_sched_switch > - __schedule > - 99.98% schedule > + 33.07% schedule_timeout > + 23.84% pipe_wait > + 20.24% do_wait > + 12.37% do_exit > + 5.34% sigsuspend > - 3.40% schedule_preempt_disabled > __mutex_lock_common.clone.8 > __mutex_lock_slowpath > - mutex_lock <---------low rate mutex blocking > + 65.71% lock_anon_vma_root.clone.24 > + 19.03% anon_vma_lock.clone.21 > + 7.14% dup_mm > + 5.36% unlink_file_vma > + 1.71% ima_file_check > + 0.64% generic_file_aio_write > - 1.07% rwsem_down_write_failed > call_rwsem_down_write_failed > exit_shm > do_exit > do_group_exit > SyS_exit_group > system_call_fastpath > - 27.61% smtpbm [kernel.kallsyms] [k] perf_trace_sched_switch > - perf_trace_sched_switch > - __schedule > - schedule > - schedule_timeout > + 100.00% sk_wait_data > + 0.46% swapper [kernel.kallsyms] [k] perf_trace_sched_switch > > > ---------------------- > 3.10-rc4 implementation (rw-sem context switch profile) > > 81.91% exim [kernel.kallsyms] [k] perf_trace_sched_switch > - perf_trace_sched_switch > - __schedule > - 99.99% schedule > - 65.26% rwsem_down_write_failed <------High write lock blocking > - call_rwsem_down_write_failed > - 79.36% lock_anon_vma_root.clone.27 > + 52.64% unlink_anon_vmas > + 47.36% anon_vma_clone > + 12.16% anon_vma_fork > + 8.00% anon_vma_free > + 11.96% schedule_timeout > + 7.66% do_exit > + 7.61% do_wait > + 5.49% pipe_wait > + 1.82% sigsuspend > 13.55% smtpbm [kernel.kallsyms] [k] perf_trace_sched_switch > - perf_trace_sched_switch > - __schedule > - schedule > - schedule_timeout > 0.11% rcu_sched [kernel.kallsyms] [k] perf_trace_sched_switch > > > Thanks. > > Tim -- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/