From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1754410Ab0ICH1j (ORCPT ); Fri, 3 Sep 2010 03:27:39 -0400 Received: from casper.infradead.org ([85.118.1.10]:35426 "EHLO casper.infradead.org" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752768Ab0ICH1i convert rfc822-to-8bit (ORCPT ); Fri, 3 Sep 2010 03:27:38 -0400 Subject: Re: [PATCHv11 2.6.36-rc2-tip 3/15] 3: uprobes: Slot allocation for Execution out of line(XOL) From: Peter Zijlstra To: Srikar Dronamraju Cc: Ingo Molnar , Steven Rostedt , Randy Dunlap , Arnaldo Carvalho de Melo , Linus Torvalds , Christoph Hellwig , Masami Hiramatsu , Oleg Nesterov , Mark Wielaard , Mathieu Desnoyers , Andrew Morton , Naren A Devaiah , Jim Keniston , Frederic Weisbecker , "Frank Ch. Eigler" , Ananth N Mavinakayanahalli , LKML , "Paul E. McKenney" In-Reply-To: <20100902174712.GA14891@linux.vnet.ibm.com> References: <20100825134117.5447.55209.sendpatchset@localhost6.localdomain6> <20100825134156.5447.43216.sendpatchset@localhost6.localdomain6> <1283415812.2059.1825.camel@laptop> <20100902174712.GA14891@linux.vnet.ibm.com> Content-Type: text/plain; charset="UTF-8" Content-Transfer-Encoding: 8BIT Date: Fri, 03 Sep 2010 09:26:17 +0200 Message-ID: <1283498777.1783.45.camel@laptop> Mime-Version: 1.0 X-Mailer: Evolution 2.28.3 Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org On Thu, 2010-09-02 at 23:17 +0530, Srikar Dronamraju wrote: > > > Current slot allocation mechanism: > > > 1. Allocate one dedicated slot per user breakpoint. Each slot is big > > > enuf to accomodate the biggest instruction for that architecture. (16 > > > bytes for x86). > > > 2. We currently allocate only one page for slots. Hence the number of > > > slots is limited to active breakpoint hits on that process. > > > 3. Bitmap to track used slots. > > > > An alternative method would be to have 1 slot per cpu, and manage the > > slot content using preemption notifiers. That gives you a fixed number > > of slots and an unlimited number of probe points. > > > > If the preemption happens to be a migration you need to rewrite the > > userspace IP to point to the new slot -- if indeed the task was inside > > one when it got preempted -- but that all should be doable. > > > > Certainly doable but it has its share of drawbacks. > 1. On every probe hit we have to copy the instruction into the > slot, so there is a performance penalty. Yeah, although I imagine its nearly free since you need to pay the cache-miss anyway. > 2 This might complicate booster probe, because the jump > instruction that follows the original instruction now actually have to > coded every time. Why can't you keep the whole replacement sequence in-tact? Simply copy it out into the slot each time. > 3. Yes migration is an issue esp > - if a thread of the same process that hit a breakpoint is scheduled into the same cpu and that newly scheduled thread hits a breakpoint. > - Something similar can happen if a multithreaded process runs on a > uniprocessor machine. -ENOPARSE ?! > 4. I dont see a need for clearing slots after post processing, but if > we need to clear we then are adding more penalties because not only are > we clearing the slots but the post processing then cant happen in > interrupt context. post-processing? you mean the probe handler? Why couldn't that be done from interrupt context? > 5. I think we are covered on the cpu hotplug too, (i.e not sure if we have > to make uprobes cpu hot plug aware.). Not if you use a slot per cpu and use preemption notifiers, the preemption notifiers will migrate the slots around. > 6. We would still be allocating a page for the slots. Unless we want > to expand to more slots than available in one page, I dont see the > disadvantages with the current approach. The current approach limits the number of probes to what fits in a page. The slot per cpu approach will have no such limit.