Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1763273Ab3ECSRw (ORCPT ); Fri, 3 May 2013 14:17:52 -0400 Received: from g6t0184.atlanta.hp.com ([15.193.32.61]:24753 "EHLO g6t0184.atlanta.hp.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1762781Ab3ECSRu (ORCPT ); Fri, 3 May 2013 14:17:50 -0400 X-Greylist: delayed 484 seconds by postgrey-1.27 at vger.kernel.org; Fri, 03 May 2013 14:17:50 EDT Subject: Re: [PATCH 1/1] cciss: bug fix to prevent cciss from loading in kdump crash kernel From: "Mike Miller (OS Dev)" Reply-To: mike.miller@hp.com, mikem@beardog.cce.hp.com To: James Bottomley Cc: Andrew Morton , Jens Axboe , LKML , LKML-scsi In-Reply-To: <1367604167.5981.40.camel@dabdike> References: <20130423172539.GA11255@beardog.cce.hp.com> <1367604167.5981.40.camel@dabdike> Content-Type: text/plain; charset="UTF-8" Date: Fri, 03 May 2013 13:09:42 -0500 Message-ID: <1367604582.11907.4798.camel@thumper.usa.hp.com> Mime-Version: 1.0 X-Mailer: Evolution 2.28.3 (2.28.3-30.el6) Content-Transfer-Encoding: 7bit Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Length: 5040 Lines: 125 On Fri, 2013-05-03 at 11:02 -0700, James Bottomley wrote: > On Tue, 2013-04-23 at 12:25 -0500, Mike Miller wrote: > > PATCH 1/1 > > > > By default the cciss driver supports all "older" HP Smart Array controllers > > and hpsa supports all controllers starting with the G6 family. There are > > module parameters that allow a user to override those defaults and use hpsa > > for any HP Smart Array controller. > > If the user does override the default behavior and uses hpsa for older > > controllers it is possible that cciss may try to load in a kdump crash > > kernel. This may happen if cciss is loaded first from the kdump initrd > > image. If cciss does load rather than hpsa and reset_devices is true we > > immediately call cciss_hard_reset_controller. This will result in a kernel > > panic and the core file cannot be created. > > This patch prevents cciss from trying to load in this scenario. > > > > Tested with 3.9.0-rc7. > > > > From: Mike > > Signed-off-by: Mike Miller > > > > --- > > drivers/block/cciss.c | 10 ++++++++++ > > 1 files changed, 10 insertions(+), 0 deletions(-) > > > > diff --git a/drivers/block/cciss.c b/drivers/block/cciss.c > > index 1c1b8e5..06c8dba 100644 > > --- a/drivers/block/cciss.c > > +++ b/drivers/block/cciss.c > > @@ -4960,6 +4960,16 @@ static int cciss_init_one(struct pci_dev *pdev, const struct pci_device_id *ent) > > ctlr_info_t *h; > > unsigned long flags; > > > > + /* > > + * By default the cciss driver is used for all older HP Smart Array > > + * controllers. There are module paramaters that allow a user to > > + * override this behavior and instead use the hpsa SCSI driver. If > > + * this is the case cciss may be loaded first from the kdump initrd > > + * image and cause a kernel panic. So if reset_devices is true and > > + * cciss_allow_hpsa is set just bail. > > + */ > > + if ((reset_devices) && (cciss_allow_hpsa == 1)) > > + return -ENODEV; > > rc = cciss_init_reset_devices(pdev); > > if (rc) { > > if (rc != -ENOTSUPP) > > Sigh, right change log, incomplete bug fix. > > Can we all agree that this is the right one? > > James I submitted 2 patches. Below the 2 are combined and make the fix complete. -- mikem > > --- > >From 746ba9f715b9037264ae0b8175c6286f5f8f62d4 Mon Sep 17 00:00:00 2001 > From: Mike Miller > Date: Thu, 18 Apr 2013 13:49:37 -0500 > Subject: [PATCH] [SCSI] cciss: bug fix to prevent cciss from loading in kdump > crash kernel > > By default the cciss driver supports all "older" HP Smart Array controllers > and hpsa supports all controllers starting with the G6 family. There are > module parameters that allow a user to override those defaults and use hpsa > for any HP Smart Array controller. > If the user does override the default behavior and uses hpsa for older > controllers it is possible that cciss may try to load in a kdump crash > kernel. This may happen if cciss is loaded first from the kdump initrd > image. If cciss does load rather than hpsa and reset_devices is true we > immediately call cciss_hard_reset_controller. This will result in a kernel > panic and the core file cannot be created. > This patch prevents cciss from trying to load in this scenario. > > Signed-off-by: Mike Miller > Cc: > Signed-off-by: James Bottomley > > diff --git a/drivers/block/cciss.c b/drivers/block/cciss.c > index 1c1b8e5..daaab88 100644 > --- a/drivers/block/cciss.c > +++ b/drivers/block/cciss.c > @@ -75,6 +75,12 @@ module_param(cciss_simple_mode, int, S_IRUGO|S_IWUSR); > MODULE_PARM_DESC(cciss_simple_mode, > "Use 'simple mode' rather than 'performant mode'"); > > +static int cciss_allow_hpsa; > +module_param(cciss_allow_hpsa, int, S_IRUGO|S_IWUSR); > +MODULE_PARM_DESC(cciss_allow_hpsa, > + "Prevent cciss driver from accessing hardware known to be " > + " supported by the hpsa driver"); > + > static DEFINE_MUTEX(cciss_mutex); > static struct proc_dir_entry *proc_cciss; > > @@ -4960,6 +4966,16 @@ static int cciss_init_one(struct pci_dev *pdev, const struct pci_device_id *ent) > ctlr_info_t *h; > unsigned long flags; > > + /* > + * By default the cciss driver is used for all older HP Smart Array > + * controllers. There are module paramaters that allow a user to > + * override this behavior and instead use the hpsa SCSI driver. If > + * this is the case cciss may be loaded first from the kdump initrd > + * image and cause a kernel panic. So if reset_devices is true and > + * cciss_allow_hpsa is set just bail. > + */ > + if ((reset_devices) && (cciss_allow_hpsa == 1)) > + return -ENODEV; > rc = cciss_init_reset_devices(pdev); > if (rc) { > if (rc != -ENOTSUPP) > > -- 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/