From mboxrd@z Thu Jan 1 00:00:00 1970 From: Keir Fraser Subject: Re: [PATCH] ACPI/APEI: fix ERST MOVE_DATA instruction implementation Date: Wed, 17 Oct 2012 14:01:54 +0100 Message-ID: References: <507EBD7002000078000A1FC9@nat28.tlf.novell.com> Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit Return-path: In-Reply-To: <507EBD7002000078000A1FC9@nat28.tlf.novell.com> List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Sender: xen-devel-bounces@lists.xen.org Errors-To: xen-devel-bounces@lists.xen.org To: Jan Beulich Cc: xen-devel List-Id: xen-devel@lists.xenproject.org On 17/10/2012 13:15, "Jan Beulich" wrote: >>>> On 17.10.12 at 14:07, Keir Fraser wrote: >> On 17/10/2012 12:53, "Jan Beulich" wrote: >> >>> Before anyone asks - yes, I have added to my todo list to put in >>> place a proper ioremap() for x86. But since I'm under some time >>> pressure right now, this was the faster route to something that >>> can be committed to see whether the regression goes away. >> >> That's good as we'll need it for >4kB per-VCPU stacks (even if we only need >> 4kB VCPU stacks, we'll want a guard page too). > > But why would you want to do that via ioremap()? Well the required missing mechanism is the same, right? The ability to allocate some unused virtual address space. We already have map_pages_to_xen(), so it's just the allocation part that is missing for either ioremap() or vcpu stacks. -- Keir > Jan >