From mboxrd@z Thu Jan  1 00:00:00 1970
Return-Path: <linux-kernel-owner@vger.kernel.org>
Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand
        id S1750947AbcIGEr7 (ORCPT <rfc822;w@1wt.eu>);
        Wed, 7 Sep 2016 00:47:59 -0400
Received: from mail-yw0-f182.google.com ([209.85.161.182]:33023 "EHLO
        mail-yw0-f182.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org
        with ESMTP id S1750791AbcIGEr6 (ORCPT
        <rfc822;linux-kernel@vger.kernel.org>);
        Wed, 7 Sep 2016 00:47:58 -0400
Date: Wed, 7 Sep 2016 10:17:52 +0530
From: Pratyush Anand <panand@redhat.com>
To: David Long <dave.long@linaro.org>,
        Catalin Marinas <catalin.marinas@arm.com>
Cc: linux-arm-kernel@lists.infradead.org, linux@arm.linux.org.uk,
        will.deacon@arm.com, Yang Shi <yang.shi@linaro.org>,
        steve.capper@linaro.org, srikar@linux.vnet.ibm.com,
        Suzuki K Poulose <suzuki.poulose@arm.com>,
        vijaya.kumar@caviumnetworks.com, linux-kernel@vger.kernel.org,
        oleg@redhat.com, Sandeepa Prabhu <sandeepa.s.prabhu@gmail.com>,
        wcohen@redhat.com, Anna-Maria Gleixner <anna-maria@linutronix.de>
Subject: Re: [PATCH 3/5] arm64: Handle TRAP_HWBRKPT for user mode as well
Message-ID: <20160907044752.GF24688@localhost.localdomain>
References: <cover.1470114993.git.panand@redhat.com>
 <ebc25cc4a2ea95cace064aa1c3a2407885623d16.1470114993.git.panand@redhat.com>
 <20160906161139.GG19605@e104818-lin.cambridge.arm.com>
 <57CF36D2.4010404@linaro.org>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <57CF36D2.4010404@linaro.org>
User-Agent: Mutt/1.6.2 (2016-07-01)
Sender: linux-kernel-owner@vger.kernel.org
List-ID: <linux-kernel.vger.kernel.org>
X-Mailing-List: linux-kernel@vger.kernel.org

On 06/09/2016:05:36:18 PM, David Long wrote:
> On 09/06/2016 12:11 PM, Catalin Marinas wrote:
> > On Tue, Aug 02, 2016 at 11:00:07AM +0530, Pratyush Anand wrote:
> > > --- a/arch/arm64/kernel/debug-monitors.c
> > > +++ b/arch/arm64/kernel/debug-monitors.c
> > > @@ -246,6 +246,8 @@ static void send_user_sigtrap(int si_code)
> > >   static int single_step_handler(unsigned long addr, unsigned int esr,
> > >   			       struct pt_regs *regs)
> > >   {
> > > +	bool handler_found = false;
> > > +
> > >   	/*
> > >   	 * If we are stepping a pending breakpoint, call the hw_breakpoint
> > >   	 * handler first.
> > > @@ -253,7 +255,14 @@ static int single_step_handler(unsigned long addr, unsigned int esr,
> > >   	if (!reinstall_suspended_bps(regs))
> > >   		return 0;
> > > 
> > > -	if (user_mode(regs)) {
> > > +#ifdef	CONFIG_KPROBES
> > > +	if (kprobe_single_step_handler(regs, esr) == DBG_HOOK_HANDLED)
> > > +		handler_found = true;
> > > +#endif
> > > +	if (!handler_found && call_step_hook(regs, esr) == DBG_HOOK_HANDLED)
> > > +		handler_found = true;
> > > +
> > > +	if (!handler_found && user_mode(regs)) {
> > >   		send_user_sigtrap(TRAP_HWBKPT);
> > 
> > Could we register kprobe_single_step_handler() via register_set_hook()
> > and only invoke call_step_hook() above?
> > 
> 
> I seem to recall a criticism of doing that in a much earlier kprobes64 patch
> of mine.  The concern was that it would cause unnecessarily more kernel
> functions to be kprobes-blacklisted.  Hence the hardcoded check and call.

Yes, all the code regions are kprobe unsafe which lie within the moment we
receive a break/single step exception to the point where it is handled for
kprobe. Therefore we must call kprobe_single_step/breakpoint_handler() before
other handlers. Otherwise, we would not be able to trace other handlers and the
functions called from those handlers.

~Pratyush