From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1753641AbbDHRIy (ORCPT ); Wed, 8 Apr 2015 13:08:54 -0400 Received: from casper.infradead.org ([85.118.1.10]:48775 "EHLO casper.infradead.org" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752605AbbDHRIu (ORCPT ); Wed, 8 Apr 2015 13:08:50 -0400 Message-Id: <20150408164813.810874878@infradead.org> User-Agent: quilt/0.61-1 Date: Wed, 08 Apr 2015 18:48:13 +0200 From: Peter Zijlstra To: mingo@kernel.org, rusty@rustcorp.com.au, mathieu.desnoyers@efficios.com, oleg@redhat.com, paulmck@linux.vnet.ibm.com, torvalds@linux-foundation.org Cc: linux-kernel@vger.kernel.org, andi@firstfloor.org, rostedt@goodmis.org, tglx@linutronix.de, peterz@infradead.org Subject: [PATCH v4 0/9] latched RB-trees and __module_address() Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org This series is aimed at making __module_address() go fast(er). The reason for doing so is that most stack unwinders use kernel_text_address() to validate each frame. Perf and ftrace (can) end up doing a lot of stack traces from performance sensitive code. On the way there it: - annotates and sanitizes module locking - introduces the latched RB-tree - employs it to make __module_address() go fast. I've build and boot tested this on x86_64 with modules and lockdep enabled. Performance numbers (below) are done with lockdep disabled. As previously mentioned; the reason for writing the latched RB-tree as generic code is mostly for clarity/documentation purposes; as there are a number of separate and non trivial bits to the complete solution. As measued on my ivb-ep system with 84 modules loaded; prior to patching the test module (below) reports: avg +- stdev Before: 1689 +- 287 [ns] per __module_address() call After: 137 +- 38 [ns] per __module_address() call Note; I have also tested things like: perf record -a -g modprobe mod_test, to make 'sure' to hit some of the more interesting paths. Changes since last time: - depend on CONFIG_PERF_EVENTS || CONFIG_TRACING -- akpm - use within_module() for jump_labels -- rusty - minor comment changes -- Compudj, mingo Rusty, please consider merging this.