From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1753106Ab2ITRgn (ORCPT ); Thu, 20 Sep 2012 13:36:43 -0400 Received: from cam-admin0.cambridge.arm.com ([217.140.96.50]:37097 "EHLO cam-admin0.cambridge.arm.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751693Ab2ITRgl (ORCPT ); Thu, 20 Sep 2012 13:36:41 -0400 Date: Thu, 20 Sep 2012 18:35:56 +0100 From: Will Deacon To: Stephen Boyd Cc: "linux-kernel@vger.kernel.org" , "linux-arm-kernel@lists.infradead.org" Subject: Re: [PATCH] ARM: hw_breakpoint: Clear breakpoints before enabling monitor mode Message-ID: <20120920173556.GQ4654@mudshark.cambridge.arm.com> References: <1348160260-19486-1-git-send-email-sboyd@codeaurora.org> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <1348160260-19486-1-git-send-email-sboyd@codeaurora.org> User-Agent: Mutt/1.5.21 (2010-09-15) Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Hi Stephen, On Thu, Sep 20, 2012 at 05:57:40PM +0100, Stephen Boyd wrote: > The reset value of the BCR, BVR, WCR, and WVR registers are all > UNKNOWN on ARMv7. Unfortunately, reset_ctrl_regs() clears these > registers *after* enabling monitor mode, not before, and so some > implementations may experience UNPREDICTABLE behavior if the > reset values of these registers are non-zero. Clear the > breakpoints before enabling monitor mode so that we don't > experience boot hangs/loops due to breakpoints being enabled > out of reset. > > Signed-off-by: Stephen Boyd Aha, thanks for the patch. We should definitely zero these registers before enabling monitor mode. However... > +/* Determine if halting mode is enabled */ > +static int halting_mode_enabled(void) > +{ > + u32 dscr; > + > + ARM_DBG_READ(c1, 0, dscr); > + > + if (WARN_ONCE(dscr & ARM_DSCR_HDBGEN, > + "halting debug mode enabled. Unable to access hardware resources.\n")) { > + return -EPERM; > + } > + return 0; > +} ...it looks like debug arch 7.1 defines this bit as UNKNOWN when the OS lock is clear, so we probably shouldn't be reading it at all. I'll pour myself a stiff drink and start reading the debug arch docs to work out what on Earth we should do. Stay tuned. Will