From mboxrd@z Thu Jan 1 00:00:00 1970 From: Ezequiel Garcia Subject: Re: [PATCH 08/28] nios2: MMU Fault handling Date: Thu, 24 Apr 2014 12:22:32 -0300 Message-ID: <20140424152232.GD1735@arch.cereza> References: <1397824031-4892-1-git-send-email-lftan@altera.com> <1397824031-4892-6-git-send-email-lftan@altera.com> <20140422143036.GD15745@arch.cereza> Mime-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: QUOTED-PRINTABLE Return-path: Content-Disposition: inline In-Reply-To: Sender: linux-kernel-owner@vger.kernel.org To: Ley Foon Tan Cc: Linux-Arch , "linux-kernel@vger.kernel.org" , "linux-doc@vger.kernel.org" , cltang@codesourcery.com List-Id: linux-arch.vger.kernel.org On Apr 24, Ley Foon Tan wrote: > On Tue, Apr 22, 2014 at 10:30 PM, Ezequiel Garcia > wrote: > > Hello Ley Foon, > > > > On Apr 18, Ley Foon Tan wrote: > >> + > >> +bad_area_nosemaphore: > >> + /* User mode accesses just cause a SIGSEGV */ > >> + if (user_mode(regs)) { > > > > I found that it's useful to add some printing here, just as ARM > > does. I carry this patch on my kernel: > > > > + printk(KERN_INFO "%s: unhandled page fault (%d) at = 0x%08lx, cause %ld\n", > > + current->comm, SIGSEGV, address, cause); > > + show_regs(regs); > > > > Do you think we could do something like it? Maybe with a compile ti= me option? > Yes, this is useful debug message. I can add this. > Prefer not to add new compile time option, I think this shouldn't > happen frequently. No, it shouldn't happen frequently. You can take a look at what ARM doe= s. They use a DEBUG_USER compile-time option, much more complex than my si= lly proposal. --=20 Ezequiel Garc=EDa, Free Electrons Embedded Linux, Kernel and Android Engineering http://free-electrons.com