Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1751798AbaDVGDW (ORCPT ); Tue, 22 Apr 2014 02:03:22 -0400 Received: from us01smtprelay-2.synopsys.com ([198.182.60.111]:52611 "EHLO smtprelay.synopsys.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1750756AbaDVGDU (ORCPT ); Tue, 22 Apr 2014 02:03:20 -0400 Message-ID: <53560613.7030801@synopsys.com> Date: Tue, 22 Apr 2014 11:32:59 +0530 From: Vineet Gupta User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:24.0) Gecko/20100101 Thunderbird/24.4.0 MIME-Version: 1.0 Newsgroups: gmane.linux.kernel.mm,gmane.linux.kernel To: Davidlohr Bueso , CC: , , , Subject: Re: [PATCH 4/6] arc: call find_vma with the mmap_sem held References: <1397960791-16320-1-git-send-email-davidlohr@hp.com> <1397960791-16320-5-git-send-email-davidlohr@hp.com> In-Reply-To: <1397960791-16320-5-git-send-email-davidlohr@hp.com> X-Enigmail-Version: 1.6 Content-Type: text/plain; charset="ISO-8859-1" Content-Transfer-Encoding: 7bit X-Originating-IP: [10.12.196.93] Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Hi, On Sunday 20 April 2014 07:56 AM, Davidlohr Bueso wrote: > Performing vma lookups without taking the mm->mmap_sem is asking > for trouble. While doing the search, the vma in question can be > modified or even removed before returning to the caller. Take the > lock (shared) in order to avoid races while iterating through > the vmacache and/or rbtree. > > This patch is completely *untested*. > > Signed-off-by: Davidlohr Bueso > Cc: Vineet Gupta > --- > arch/arc/kernel/troubleshoot.c | 7 ++++--- > 1 file changed, 4 insertions(+), 3 deletions(-) > > diff --git a/arch/arc/kernel/troubleshoot.c b/arch/arc/kernel/troubleshoot.c > index 73a7450..3a5a5c1 100644 > --- a/arch/arc/kernel/troubleshoot.c > +++ b/arch/arc/kernel/troubleshoot.c > @@ -90,7 +90,7 @@ static void show_faulting_vma(unsigned long address, char *buf) > /* can't use print_vma_addr() yet as it doesn't check for > * non-inclusive vma > */ > - > + down_read(¤t->active_mm->mmap_sem); Actually avoiding the lock here was intentional - atleast in the past, in case of a crash from mmap_region() - due to our custom mmap syscall handler (not in mainline) it would cause a double lockup. However given that this code is now only called for user contexts [if user_mode(regs)] above becomes moot point anyways and it would be safe to do that. So, yes this looks good. A minor suggestion though - can you please use a tmp for current->active_mm as there are 3 users now in the function. Acked-by: Vineet Gupta Thx -Vineet > vma = find_vma(current->active_mm, address); > > /* check against the find_vma( ) behaviour which returns the next VMA > @@ -110,9 +110,10 @@ static void show_faulting_vma(unsigned long address, char *buf) > vma->vm_start < TASK_UNMAPPED_BASE ? > address : address - vma->vm_start, > nm, vma->vm_start, vma->vm_end); > - } else { > + } else > pr_info(" @No matching VMA found\n"); > - } > + > + up_read(¤t->active_mm->mmap_sem); > } > > static void show_ecr_verbose(struct pt_regs *regs) > -- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/