From mboxrd@z Thu Jan 1 00:00:00 1970 From: Horms Date: Wed, 30 Aug 2006 08:27:19 +0000 Subject: Re: IA64 kexec/kdump 2.6.18-rc5 patch Message-Id: <20060830082717.GA21121@verge.net.au> List-Id: References: <1156837594.2598.15.camel@linux-znh> In-Reply-To: <1156837594.2598.15.camel@linux-znh> MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit To: linux-ia64@vger.kernel.org On 30 Aug 2006 06:03:23 +0800, Zou Nan hai wrote: > On Wed, 2006-08-30 at 03:38, Bjorn Helgaas wrote: >> On Tuesday 29 August 2006 01:46, Zou Nan hai wrote: >> > +#ifdef CONFIG_KEXEC >> > +void >> > +ioc_iova_disable(void) >> > +{ >> >> Ugh. If you really need this functionality (which I have to say >> looks >> like a band-aid), it probably should be a platform vector. And >> should >> be split into a separate patch. >> > Hi Bjorn, > The ioc_iova_disable code comes from Aziz in HP, I have almost > no idea > of how IOMMU works on HP platform. > I am looking for an HP machine with IOMMU to test. That sounds like even more reason to break it out into a separate patch. Actually, I really think that you sould either provide a set or smaller patches, or incremental patches. Its quite hard to follow what is changing with the current jumbo-patch format. Its also quite unclear what if any portions you would like merged, and even which tree you want them merged into - you patches are against 2.6.18-rc5, but the ia64 tree seems the most likely path, and your code doesn't apply there because some of the changes are already present. -- Horms H: http://www.vergenet.net/~horms/ W: http://www.valinux.co.jp/en/